Historic, archived document
Do not assume content reflects current
scientific knowledge, policies, or practices
IssUed June 29, 190?.
U. S. DEPARTMENT OF AGRICULTURE.
BUREAU OF CHEMISTRY— BULLETIN No. 105.
H. W WILEY, Chief of Bureau.
PROCEEDINGS
TWENTY-THIRD ANNUAL. CONVENTION
Association of Official Agricultural Chemists,
WASHINGTON, D. C, NOVEMBER 14-16, 1906.
EDITKD BY
HARVEY W. WILEY,
Secretary of the Association.
WASHINGTON:
GOVERNMENT PRINTING OFFICE
1907.
I
Issued June 29, 1907.
U. S. DEPARTMENT OF AGRICULTURE,
BUREAU OF CHEMISTRY— BULLETIN No. 105.
H. W. WILEY, Chief of Bureau.
PROCEEDINGS
TWENTY-THIRD ANNUAL CONVENTION
Association of Official Agricultural Chemists.
WASHINGTON, D. C, NOVEMBER 14-16, 1906.
EDITED BY
HARVEY W. WILEY,
Secretary of the Association.
WASHINGTON:
government PRINTING OFFICE,
1907.
LETTER OF TRANSMITTAL.
U. S. Department of Agriculture,
Bureau of Chemistry,
Washington, D. C, Fehruary 28, 1907.
Sir: 1 have the honor to transmit for your approval the proceed-
ings of the Twenty-third Annual Convention of the Association of
Official Agricultural Chemists. The growth in the work of the asso-
ciation has made it imperative that both papers and discussion be
condensed as far as possible, and all methods or other material
which can be found in other generally available sources are given
only by reference. I recommend that tins report be published as
Bulletin No. 105 of the Bureau of Chemistry.
Respectfully,
H. W. Wiley,
Chief of Bureau.
Hon. James Wilson,
Secretary of Agriculture.
(3)
CONTENTS.
Wednesday — Morning Session.
Page.
Members and visitors present 7
Food adulteration 11
Report on colors. By E. F. Ladd 11
Experimental work on coloring matters. By H. M. Loomis 12
The detection of certain commercial colors alleged to be vegetable colors. By
A. G. Nickles •- 12
Report on saccharine products. By C. H. Jones 14
Report on fruit products. By Hermann C. Lythgoe 19
Report on beer. By H. E. Barnard 20
Report on distilled liquors. By C. A. Crampton 20
Fuller's earth test for caramel in vinegar. By W. L. Dubois 23
Report on flavoring extracts. By E. M, Chace 25
Report on baking powders. By W. M. Allen 28
Report on fats and oils. By L. M. Tolman 29
Report on dairy products. By Albert E. Leach 37
Report on condiments other than spices. By R. E. Doolittle 39
Report on tea and coffee. By C. D. Howard 41
Review of methods for analysis of tea. By R. E. Doolittle and F. O. Wood-
ruff 46
Report on food preservatives. By W, L. Dubois 51
Wednesday — Afternoon Session.
Report on determination of water in foods. By F. C. Weber 58
Report on cereal products. By A. McGill 66
Report of Committee C on recommendations of referees. By H. C. Lythgoe. . . 73
Report on nitrogen. By James H. Gibboney 76
The detection of peat in commercial fertilizers. By John Phillips Street 83
Comparative work on nitrogen estimations by the Kjeldahl and Gunning
methods and by a combination of the two methods. By Thomas S. Gladding. . 85
Report on the separation of nitrogenous bodies in cheese. By R. Harcourt 86
Report on the separation of vegetable proteids. By Harry Snyder 88
Thursday — Morning Session.
Appointment of committees 91
Report on the separation of meat proteids. By F. C. Cook 91
Report on dairy products. By F. W. WoU 98
Subreport on analysis of dairy products. By G. E. Patrick and M. Boyle 106
Fat determinations in cheese by the Gottlieb and the ether-extraction methods.
By George A. Olson 109
Determination of the acidity of cheese, . By Alfred W. Bosworth 110
Report on foods and feeding stuffs. By J. K. Haywood 112
Report on sugar. By C. A. Browne and J. E. Halligan 116
Report on chemical methods of sugar analysis. By L. S. Munson 125
(5)
Thursday- — Afternoon Session.
Page.
The chemical determination of sulphites in sugar products. By W. D. Home. 125
Report on medicinal plants and drugs. By L. F. Kebler 127
Report on soils. By J. H. Pettit 142
Determination of total potassium in soils. By J. H. Pettit and A. Ystgard . . . 147
Report of committee on revision of methods. By J. K. Haywood 148
Report of committee on nominations. By C. D. Woods 150
Friday — Morning Session.
Report on inorganic plant constituents. By W. W. Skinner 151
Report of Committee B on recommendations of referees. By E. B. Holland 154
Report on phosphoric acid. By B. W. Kilgore 157
Report on determination of iron and alumina in phosphates and on the citrate
solution. By J. M. McCandless 157
Address by Assistant Secretary of Agriculture .* 161
Report On tannin. By H, C. Reed 161
Determination of kerosene in kerosene emulsion by the centrifugal method. By
G. E. Colby 165
Methods of analysis of lead arsenate. By J. K, Haywood 165
Report of the committee on food standards. By Wm. Frear 168
Report of the committee on fertilizer legislation. By H. W. Wiley 174
Adulteration of commercial fertilizers. By J. M. McCandless 177
Report of the committee on definition of plant food. By H. W. Wiley 178
Friday — Afternoon Session.
Report of the committee on the testing of chemical reagents. By L. F. Kebler. 181
Address by Dr. Alexius de 'Sigmond 188
Report on potash. By A. L. Knisely 190
Report of Committee A on recommendations of referees. By R, J. Davidson 196
Report of committee on the president's address. By F. W. WoU 199
Report of committee on unification of terms for reporting analytical results. By
R. J. Davidson 200
Report of the committee on resolutions. By L. L. Van Slyke 200
Officers, referees, and committees of the Association of Official Agricultural
Chemists, 1907 201
Constitution of the Association of Official Agricultural Chemists 204
Index 207
ILLUSTRATION.
Page.
Fig. 1. Comparison of refractometric examination of milk serum and the deter-
mination of solids not fat for detection of added water 38
PROCEEDINGS OF THE TWENTY-THIRD ANNUAL CONVENTION
OF THE ASSOCIATION OF OFFICIAL AGRI-
CULTURAL CHEMISTS.
KIRST DAY.
WEDNESDAY— MORNING SESSION.
The twenty-third annual convention of the Association of Official
Agricultural Chemists was called to order by the president, Mr. C. G.
Hopkins, of Urbana, 111., at 10 o'clock on the morning of November 17,
at the George Washington University, Washington, D. C.
The following members and visitors registered during the
convention :
MEMBERS AND VISITORS PRESENT.
Adams, Arthur B., Bureau of Internal Revenue, U. S. Treasury Department, Wash-
ington, D. C.
Albrech, Maximilian C, Bureau of Chemistry, U. S. Department of Agriculture,
Washington, D. C.
Allen, William M., Department of Agriculture, Raleigh, N. C.
Alwood, William B., Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Ames, John W., Agricultural Experiment Station, Wooster, Ohio.
Arnold, Robert B., Box 726, Richmond, Va.
Bartlett, James M., Agricultural Experiment Station, Orono, Me.
Beal, W. H., Office of Experiment Stations, U. S. Department of Agriculture, Wash-
ington, D. C.
Bell, James M., Bureau of Soils, U. S. Department of Agriculture, Washington, D. C.
Bigelow, W. D., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Bishop, Harry E., Indiana Laboratory of Hygiene, Indianapolis, Ind.
Bizzell, James A., Agricultural Experiment Station, Ithaca, N. Y.
Blair, A. W., Agricultural Experiment Station, Lake City, Fla.
Bowker, W. H., Bowker Fertilizer Co., Boston, Mass.
Brinton, Clement S., Food Laboratory, U. S. Department of Agriculture, Philadel-
phia, Pa.
Browne, Charles A., jr., Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Burnet, Wallace C, Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Burton, Joseph Q., Atlanta, Ga.
Cameron, Frank K., Bureau of Soils, U. S. Department of Agriculture, Washington,
D. C.
(7)
8
Carpenter, F. B., Virginia-Carolina Chemical Company, Richmond, Va.
Carroll, John S., German Kali Works, Atlanta, Ga.
Chace, Ed. MacKay, Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Chamberlin, Joseph S., Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Church, C. G., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Cochran, Carlos B., Department of Agriculture, Westchester, Pa. ■
Cook, Frank C, Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Crampton, Charles A., Bureau of Internal Revenue, U. S. Treasury Department,
Washington, D. C.
Davidson, Robert J., Agricultural Experiment Station, Blacksburg, Va.
Donk, Marion G., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C. >
Doolittle, Roscoe E., Food Laboratory, U. S. Department of Agriculture, New York,
N. Y.
Dox, Arthur W., Agricultural Experiment Station, Storrs, Conn.
Doyle, Aida M., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Dubois, Wilbur L., Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
EUett, Walter B., Agricultural Experiment Station, Blacksburg, Va.
Frear, William F., State College, Pa.
Frost, Howard V., Food Laboratory, U. S. Department of Agriculture, Chicago, 111.
Fuller, F. D., Department of Agriculture, Harrisburg, Pa.
Gamble, Wm. P., Ontario Agricultural College, Guelph, Canada.
Gascoyne, Wm. J., 2741 North Charles street, Baltimore, Md.
Given, Arthur, Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Gooden, Edward H., Bureau of Internal Revenue, U. S. Treasury Department, Wash-
ington, D. C.
Goodrich, Charles E., Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Gore, H. C, Bm'eau of Chemistry, U. S. Department of Agriculture, Washington, D. C.
Goss, Willard L., Bureau of Plant Industry, U. S. Department of Agriculture, Wash-
ington, D. C.
Gould, Ralph A., Food Laboratory, College of Agriculture, Berkeley, Cal.
Graham, J. J. T., Agricultural Experiment Station, College Park, Md.
Gray, C. Earl, Bureau of Animal Industry-, U. S. Department of Agriculture, Washing-
ton, D. C.
Hand, William F., Agricultural Experiment Station, Agricultural College, Miss.
Hanson, Herman H., Agricultural Experiment Station, Orono, Me.
Hardin, Mark B., Agricultural Experiment Station, Clemson College, S. C.
Hargrove, Julian O., 1603 O street, Washington, D. C.
Harris, C. D., Assistant State Chemist, Raleigh, N. C.
Harrison, Channing W., Food Laboratory, U. S. Department of Agriculture, New
Orleans, La.
Hart, Benj. R., Agricultiual Experiment Station, Lexington, Ky.
Hartwell, Burt L., Agricultural Experiment Station, Kingston, R. I.
Haywood, John K., Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Herff, B. von, 93 Nassau street, New York, N. Y.
Hill, D. H., Bureau of Chemistry, U. S. Department of Agriculture, Washington, D. C.
Hite, B. H., Agricultural Experiment Station, Morgantown, W. Va.
Holland, Ed. B., Hatch Experiment Station, Amherst, Mass.
Holland, Herbert J., Food Laboratory, College of Agriculture, Berkeley, Cal.
Hoover, Geo. B., Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Hopkins, Cyiil G., Agricultural Experiment Station, Urbana, 111.
Home, Wm. D., National Sugar Refining Company, Yonkers, N. Y.
Houghton, Harry W., Bureau of Chemistry, U. S. Department of Agiiculture, Wash-
ington, D C.
Howard, Burton J., Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Howard, Charles D., Board of Health, Concord, N. H.
Jones, Charles H., Agricultural Experiment Station, Burlington, Vt.
Jones, Wm. J., jr.. Agricultural Experiment Station, Lafayette, Ind.
Kebler, Lyman F., Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Keister, John T., Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Knight, Howard L., Office of Experiment Stations, U. S. Department of Agriculture,
Washington, D. C.
La Bach, James 0., Agricultural Experiment Station, Lexington, Ky.
Langley, Clifford, 64 Irving place. New York, N. Y.
Law, Leroy M., Bureau of Internal Revenue, IT. S. Treasury Dex^artraent, Washington,
D. C.
Lawson, H. W., Office of Experiment Stations, U. S. Department of Agriculture,
Washington, D. C.
Leavitt, Sherman, Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Le Clerc, J. A., Bureau of Chemistry, IT. S. Department of Agriculture, Washington,
D. C.
Lind, Herman, Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Langworthy, C. F., Office of Experiment Stations, U. S. Department of Agriculture,
Washington, D. C.
Lythgoe, Herman C, State Board of Health, Boston, Mass.
McCandless, Jno. M., State Chemist, Atlanta, Ga.
McClelland, Byron, Bureau of Chemistry. U. S. Department of Agriculture, Washing-
ton, D. C.
McDonnell, Henry B., Agricultural Experiment Station, College Park, Md.
McGill, Anthony, Chemist to Canadian Government, Ottawa, Canada.
Magruder, E. W., Department of Agriculture, Richmond, Va.
Mason, Glen F., H. J. Heinz Company, Pittsburg, Pa.
Mathis, Walter R., Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Mewborne, Robert G., Kentucky Tobacco Product Company, Louisville, Ky.
Mooers, Charles A., Agricultural Experiment Station, Knoxville, Tenn.
Moore, C. C, Bureau of Chemistry, IT. S. Department of Agriculture, Washington,
D. C.
Morgan, Jerome J., Agricultural Experiment Station, College Park, Md.
Munson, L. S., The Ault Wibourg Co., Cincinnati, 0.
10
Outwater, Raymond, Agricultural Experiment Station, College Park, Md.
Palmore, Julian I., Assistant State Chemist, College Park, Md.
Parker, Charles E., Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Patrick, Geo. E., Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Patten, Harrison E., Bureau of Soils, U. S. Department of Agriculture, Washington,
D. C.
Patterson, H. J., Agricultui'al Experiment Station, College Park, Md.
Penny, Charles L., Agricultural Experiment Station, Newark, Del.
Peter, Alfred M., Agricultural Experiment Station, Lexington, Ky.
Pettit, James H., Agricultural Experiment Station, Urbana, 111.
Price, Thomas M., Bureau of Animal Industry, U. S. Department of Agriculture,
Washington, D. C.
Prit chard, Wm. P., City Bacteriologist, Fall River, Mass.
Reed, Howard S., Bureau of Soils, U. S. Department of Agriculture, Washington,
D. C.
Robb, John B., Department of Agriculture, Richmond, Va.
Robinson, Floyd W., State Analyst, Lansing, Mich.
Ross, B. B., State Chemist, Auburn, Ala.
Runyan, Elmer G., Inspector of Gas and Meters, Washington, D. C.
Schreiber, Herman, Bureau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Schreiner, Oswald, Bureau of Soils, U. S. Department of Agriculture, Washington,
D. C.
Schulz, Henry L., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Seidell, Atherton, Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Sherwood, Sidney Forsythe, Bureau of Chemistry, IT. S. Department of Agriculture,
Washington, D. C.
'Sigmond, Alexius A. J. de. University of Budapest, Hungary.
Sindall, Harry E., Bureau of Chemistry, IJ. S. Department of Agriculture, Washing-
ton, D. C.
Skinner, W. W., Bureau of Chemistry, IT. S. Department of Agriculture, Washington,
D. C.
Smith, Bernard H., Food Laboratory, U. S. Department of Agriculture, Boston, Mass.
Smith, Edwin, jr., George Washington University, Washington, D. C.
Smoot, Charles C, Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Stiles, Geo. W., jr., Bm-eau of Chemistry, U. S. Department of Agriculture, Washing-
ton, D. C.
Stillwell, Albert G., 36 Gold street. New York, N. Y.
Straughn, M. N., Agricultural Experiment Station, College Park, Md.
Street, John Phillips, Agricultural Experiment Station, New Brunswick, N. J.
Sullivan, Arthur L, Bureau of Internal Revenue, U. S. Treasury Department, Wash-
ington, D. C.
Syme, Wm. A. Agricultural Experiment Station, Raleigh, N. C.
Taber, Walter C, Bureau of Soils, V. S. Department of Agriculture, Washington, D. C.
Tolman, L. M., Bureau of Internal Revenue, U. S. Treasury Department, Washington,
D. C.
Trescot, T. C, Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
11
Van Slyke, L. L., Agricultural Experiment Station, Cieneva, N. Y.
Veitch, F. P., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Vivian, Alfred, State University, Columbus, Ohio.
Walker, Percy H. , Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D.C.
Warner, H. J., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Weber, F. C, Bureau of Chemistry, U. S. Department of Agriculture. W^ashington,
D. C.
Wesson, David, 24 Broad street, New York, N. Y.
White, Horace I>., State Board of Health, Burlington, Vt.
Wiley, H. W., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D. C.
Wiley, Raymond C, Agricultural Experiment Station, College Park, Md.
Wilson, Clarence P., Bureau of Chemistry, U. S. Department of Agriculture, Wash-
ington, D. C.
Winton, A. L., Agricultural Experiment Station, New Haven, Conn.
Woll, Fritz W., Agricultural Experiment Station, Madison, Wis.
Woods, Charles D., Agricultural Experiment Station, Orono, Me.
Young, Wm. J., Bureau of Chemistry, U. S. Department of Agriculture, Washington,
D.C.
POOD ADULTEEATIOK
No general report on food adulteration was submitted and the
reports of the associate referees were received in the following order:
EEPOET ON OOLOES.
By E. F. Ladd, Associate Referee.
In view of the fact that there is little definite knowledge with regard to color reac-
tions, and especially as to the behavior of natural colors with reagents, it seemed
desirable this year that some special attention be given to this phase of the work.
Unfortunately this investigation must be conducted in laboratories already over-
crowded with other lines of work, and several of the chemists cooperating have been
unal)le to carry their experiments far enough to make a definite report. The present
paper, therefore, may be said to be only a preliminary report on the work undertaken.
The investigation has consisted largely of individual work on special phases rather
than cooperative work on samples sent out. But two reports have been received—^
one dealing with solubility and extraction of colors and the other with the detection
of certain coinmercial colors designated as vegetable colors and used in food products.
The reports expected on colors containing tin, antimony, etc., and on the testing of
certain colors not readily extracted from an acid solution by means of amyl alcohol,
are not yet ready for presentation.
At times it is with difficulty that the coloring matter is extracted from preserved
fruits by means of amyl alcohol. It has been suggested (Pharmaceutisch Weekblad)
that the use of ethyl alcohol and amyl alcohol in equal volumes as the extracting
medium will prove more satisfactory than the use of amyl alcohol alone. The test
is made as follows: Five grams of material are rubbed up with 5 cc of water and this
warmed with 20 grams of ethyl and methyl alcohol in concentrated form. After
cooling, additional water is added and if foreign coloring matters are present the same
12
will be contained in the layer of alcohol, which readily separates from the rest ot the
mixture. Xot enough work has been done to test the value of this method of
extraction.
Eecommexdatioxs.
I repeat the suggestions of the associate referee for 1905, namely, that the members
of the association and collaborators interested in the subject of colors do what original
work their time will permit on the points suggested below:
(1) Solubility of the coal-tar and vegetable dyes in various solvents (^ ether, acetic
ether, petroleum ether, methyl and ethyl alcohols, acetone, etc.), aiTanged accord-
ing to theii" solubility — as. easily soluble, difficultly soluble, and insoluble.
(2) Extractive valties of the various solvents for dyes in neutral, acid, and alkaline
solutions.
(3) Characteristics of the coloring matters as contained in fresh fruits, vegetables,
wines, etc., with reagents and solvents, and their respective dyeing properties.
~ (4) Testing such new schemes as may appear in the various current chemical jour-
nals, and such as have appeared during the last few years.'
EXPEEIMENTAL WOEK ON OOLOEING MATTEES.
By H. M. Loo^ns.
Tlie siibreport on colors by ^Ir. H. M. Loomis, chemist, PennsTl-
vania Department of Agriculture, is piiblislied as Circular Xo. 35 of
the Bureau of Chemistry, being too lengthy for pttbhcation in the
Proceedings. This report includes four tables on the follo^ving sub-
jects: Solubilities of colors: extraction of colors: color reaction on
dyed fibers, and reactions of colors in aqueotis soltition and with
concentrated sttlphuric acid.
The stibreport on colors by ^Ir, Xickles follows:
TKE DETECTION OF CEETAIN COMMEEOIAL C0L0E8 ALLEGED TO BE
VEGETABLE OOLOES.
By A. G. XiCEXES.
The investigation of colors during the past year in the laboratory- of the Xorth
Dakota Agricultural College consisted in an effort to find some general method by
which the commercial colors alleged to be vegetable dyes which are used in foods
and beverages on the market to-day can be distinguished from colors of coal-tar origm.
A line of such commercial colors from two of the leading manufacturing firms was
used as a basis for the work. The plan of the investigation included the extraction
of colors from their solution by immiscible solvents, the fixing of colors on wool,
double dyeing method, and color reactions in aqueous solutions with certain reagents.
The accompanying table shows that thirteen of the fourteen colors examined gave
a reaction for coal-tar dye, according to the double dyeing method of Sostegni and
Carpentieri. Many of these alleged vegetable dyes which have been sent in for
examination during the year have been fotmd to give a similar permanency of color.
In testing such of these dyes as color wool in the second dyeing, make carefid note of
the following points:
I'se as little of the dye as possible: if the color is taken up readily in the first bath,
notice carefully the reactions as the wool is boiled in the second acid bath. In nearly
13
every instance, if the color is of the class under consideration the solution will become
more or less colored, while if the color is a coal-tar dye the solution will remain per-
fectly clear. There is a lack of brilliancy in all the alleged vegetable colors that dye
wool, while on the other hand this feature is well marked in the coal-tar colors.
Since complete reliance can not be placed on the use of the Sostegni and Carpentieri
method for the identification of coal tar colors, we were led to investigate the following
method:
Extract the colors by the use of solvents such as ether, amyl alcohol, acetic ether,
and acetone from solutions both in alkaline and acid condition. Evaporate the
combined extracts separately to dryness in the presence of a piece of wool, and then
make tests on the dyed fiber. Comprehensive tables of color and their reactions will
enable the analyst to classify, if not identify, the color by the use of this method.
Most vegetable dyes give specific color reactions with ammonia. This fact has
prompted quite a thorough investigation of the differentiation of the alleged vegetable
colors from those of coal tar origin. The writer has observed the following reactions
in working with the former class of commercial colors:
Place about 100 cc of water in a beaker of 150 cc capacity, add just enough of the
dye to color the solution, avoiding too deep a color. Next add ammonia water (sp. gr.
0.90) in excess, from 30 to 40 cc, stir well, and allow to remain 12 or 14 hours and
note changes. Some colors, however, change more rapidly than others. Blue colors
become colorless. Of the red colors some will change to a much deeper tinge, while
in the case of cudbear the color is changed to violet. The green, yellow, orange,
and brown colors are changed to different shades of amber. With the coal-tar dyes
there is a permanency of color not met with in the alleged vegetable dyes. There
are, however, some coal-tar dyes that give a color reaction with ammonia, but by
making the solution acid the original color is restored.
In the examination of a color by the use of the Sostegni and Carpentieri double
dyeing method and then carefully observing the color reactions with ammonia, the
analyst can quite easily distinguish between coal tar and the alleged vegetable colors.
The following table summarizes the results obtained:
Reaction of commercial colors, said to be of vegetable origin, with different reagents.
Character
of dye-
stuff.
Ether extract f rom—
Amyl alcohol extract from—
CoTTiTnercial color.
Acid
solution.
Ammoniacal
solution.
Acid
solution.
\mmoniacaI
solution.
Carmine red
Powder. . .
do
Colorless
do
Colorless
do
Red
Colorless
do
Do.
Vegetable red
.do
do
.do
...do
Do.
Do
Paste
do
do
do
Do.
....do...
do
do
Blue
Do.
Vegetable yellow
do
do
. .do
Yellow
do
Do.
Do
Powder. . .
do
do
Do.
Paste
do....
Powder. . .
Orange
Colorless
do
Orange
Colorless
do
Orange
Light green...
do
Orange
Colorless.
Do.
Vegetable green
Do
Paste. .
...do
..do
Colorless
do
Do.
Vegetable blue .
do
do
do
Do.
Do
Powder. . .
do
do
do
Do.
Paste
do
do
Brown
Do.
14
Reaction of commercial colors, said to he of vegetable origin, etc. — Continued.
i
Character
of dve-
stuff.
Acetic ether extract from—
Reaction with
ammonium hy-
droxid in aque-
ous solution
after standing
14 hours.
Commercial color.
Acid
solution.
Ammoniacal
solution.
Remarks.
Carmine red
j Powder. . .
■ do
do
Paste
Slightly col-
ored.
Red
do
do
Colorless
do
do
YeUow
Colorless
do
Deep orange . .
Colorless
do
do
do
do
do
Dark red
Cladonol red
Violet .
in second dye-
ing (Sostegni
& Carpentieri
method).
Colors wool in
Vegetable red
do
second dyeing.
Do.
Do....:
.. ..do
Do.
Vegetable violet . . .
do
do
Powder.
Colorless
do
do
Pinkish
Do.
Vegetable yeUow. . .
Amber
Do.
Do
do
Do.
Vegetable orange . .
Vegetable green
i Paste
1 do
Powder...
Orange
Colorless
do
Very light amber.
Amber
Do.
Do.
Do
do.,
Do.
Lazulem blue
Paste.
do
Becomes colorless,
do
Do.
Vegetable blue
do
do
do
Do.
Do
do
Do.
Vegetable brown. . .
Paste
Brown
Dark amber
Do.
EEPOET 0'^ SAOOHAEINE PEODUCTS.
C. H. JoxES, Associate Referee.
The work of the referee on saccharine products has been confined to maple sugar
and sirup, and particular attention has been given to malic acid value. The method
sent out this year was slightly modified from that reported at the last meeting of
this association. a Such changes as were made, however, were based on data obtained
by Mr. Julias Hortvet and the writer, who conducted careful tests to locate if possible
the cause of the variation in results reported in 1905.
Mr. Hortvet's correction applies to the method of washing the precipitated calcium
malate. Instead of attempting to "wash until free from soluble calcium salt," the
washing should cease when there is practical freedom from chlorids. The following
figures furnished by Mr. Hortvet illustrate this:
Table 1. — Effect of washing on malic acid lalue {Hortvet).
[Precipitated calcium malate in alcohol, washed with hot 75 per cent alcohol.]
No.
Filtrate and
washings.
0.50 per cent
malic acid
solution in
sirup.
Vennont
sirup.
Ohio
sirup.
1
2
3
4
5
6
cc
a200
6 125
&105
100
96
94
0.34 1.
.38 .
.51 i-
.58
.62 1
.69
0.43
.47
0.54
.59
.61
aNea
rly.
6 Approximately.
No. 1. Washed until practically no reaction for calcium. It appears that calcium
malate is slightly soluble in the hot 75 per cent alcohol.
Xo. 2. Washed beyond the point at which silver nitrate ceased to give a precipitate
insoluble in nitric acid. There was. howeA'-er, a precipitate with silver nitrate soluble
in nitric acid. Slight reaction for calcium.
oU. S. Dept. Agr., Bm-eau of Chemistry, Cir. 23, 1905.
15
No. 3. Slightly lieyond the point at which silver nitrate gave a prec'pitate insoluble
in nitric acid.
No. 4. Precipitate with silver nitrate showed only faint milkiness when treated with
strong nitric acid.
Nos. 5 and 6. More chlorids shown in last drops of filtrate than in No. 4.
Your referee, after conducting a series of tests on the reagents called for in the deter-
mination of malic acid value, concluded that the amount of ammonia added to make
the solution "slightly alkaline^' was a disturbing factor, and quite largely responsible
for some variations and many of the high results previously reported. It was learned
from correspondence with different chemists familiar with the test that from a drop to
one or more cubic centimeters of ammonia had been used.
Table 2. — Test on reagents used for obtaining malic acid value (Jones).
No.
Calcium
chloricl.
Water.
Ammonia.
Alcohol.
Malic acid
value.
cc
cc
cc
1
1
15
1 drop.
60
0.14
2
1
15
2 drops.
60
.26
3
1
15
0.5 cc.
60
.51
4
1
15
1 cc.
60
.78
5
2
15
0.5 cc.
60
.57
6
1
15
0
60
.00
7
1
15
0
60
.015
8
1
15
0
60
.02
The malic acid value figure is seen to vary directly with the excess of ammonia
present, due to the precipitation of lime after heating and standing overnight.
Three samples were prepared for cooperative work. No. 1 was a sugar containing
60 per cent cane and 40 per cent maple sugar, No. 2 contained 60 per cent of light
brown sugar « and 40 per cent of maple, while No. 3 was pure maple sugar. All of the
samples were sugared off at 238° F.
Eighteen sets of samples were sent to chemists who had requested the same, together
with the following methods of analysis and a letter of transmittal:
METHODS OP ANALYSIS.
Preparation of savijAe.
Just previous to analysis, pulverize or shave each cake, mix, and transfer to small,
wide-mouth stoppered bottles.
Ash.
AVeigh 5 grams of sugar into a tared platinum dish; heat over asbestos board until
the contents are thoroughly carbonized ; transfer to a muffle and burn at low red heat
to a white or grey ash. Cool in a desiccator and weigh quickly. Dissolve the residue
in about 40 cc of hot water and boil gently tor two minutes, using care to avoid spatter-
ing. Filter through a small ashless filter and wash w4th hot water until the filtrate
amounts to about 100 cc. Transfer filter and contents to the original platinum dish
and incinerate at low red heat as before. Cool and weigh. Cool and titrate the solu-
tion containing the soluble ash with tenth-normal hydrochloric acid, using methyl
orange as indicator. Add to the insoluble ash an excess of the tenth-normal acid
(10 cc is usually sufficient) and about 30 cc of water. Boil gently until solution is
complete. Cool and titrate wnth tenth-normal sodium hydroxid, using methyl orange
as indicator. Calculate from the results obtained the percents of total ash, water
soluble ash, insoluble ash, and the alkalinity of the soluble and insoluble ash, expressed
as cubic centimeters of tenth-normal hydrochloric acid for ash of 1 gram of sample.
^Analysis: Ash, 0.53 per cent; soluble ash, 0.47 per cent; insoluble ash, 0.06 per
cent; alkalinity, soluble ash, 0.50 cc; alkalinity, insoluble ash, 0.10 cc; malic acid
value, 0.02 per cent.
16
Malic acid value.
Weigh 6y%- grams of the sample into a 200 cc beaker and. add water to make a volume
of 20 cc. Add 1 cc of a 10 per cent solution of calciiun chlorid and heat to boiling.
Now add 60 cc of 95 per cent alcohol, cover the beaker with a watch glass and heat for
one-half horn- on a water bath. Remove and let stand over night. Filter (through
a 9 cm Xo. 589 S. & S. filter) by decantation. Transfer precipitate to the filter by
washing with hot 75 per cent alcohol and continue the washing until entire filtrate
measures 100 cc. Dry and ignite. Add from 15 to 20 cc of tenth-normal hydrochloric
acid to the ignited residue, dissolve by careful boiling, cool and titrate the excess of
acid with tenth-normal sodium hydroxid. using methyl orange as indicator. One-
tenth of the number of cubic centimeters of acid neutralized expresses the result.
Report results as follows:
1. Total ash, per cent.
2. Soluble ash, per cent.
3. Insoluble ash, per cent.
4. Alkalinity of soluble ash. as cubic centimeters required for 1 gram of material.
5. Alkalinity of insoluble ash. as cubic centimeters required for 1 gram of material.
6. Malic acid value.
7. Comments and suggestions.
Thirteen analysts, representing twelve laboratories, have furnished results shown
in the following table:
Table 3. — Results of cooperative maple sugar uorJ:.
No. 1.— ADULTEEATED. 60 PER CENT CANE, 40 PER CENT MAPLE.
Analyst.
Total
ash.
Soluble.
ash.
Insoluble
ash.
Alkalin- ; AlkaUn- ;
ity of ity of
soluble insoluble
ash. ash.
Mahc
acid
value.
\Per
A. T. Charron. Otta^va, Canada
E. B. Holland. Amherst, Mass !
C. P. Moat, Burlington, Vt '
M. C. Albrech, Washington, D . C J
E. Monroe Bailey, New Haven, Conn <
Vr.B. Pope. Concord. X. H
A. P. Sy, Bufialo, X. Y ■
C. H. Jones, Burlington. Vt '
A. G. Xiekles, Agricultural College, X.Dak
A. Vahn, Ottawa, Canada <
J. A. MiUer, Buffalo. X. Y
R. M. West. St. Paul. Minn
F. T. Shutt, Ottawa, Canada
Average
cent.
0.29
.30
.32
.30
.33
.38
.37
.33
.35
.32
.29
.33
.35
.31
cent.
0.15
22
.20 i
.22 I
.28 i
.26 !
.25
.19
.26
.20
.18
.23
.23
.20
.21
Per cent.
0.14
.12
cc.
0.29
.29
.30
.12
.12
.20
.19
.30
1.18
.29
.21
cc.
0.43
.34
.26
.22
.24
.30
.28
.40
.32
.28
a. 20
.25
.26
.40
.43
.13
.12
.03
.03
.08
.03
.11
a. 01
.09
.10
.08
.09
.33 I
.22
.11
.24
.31
Xo. 2.— ADULTERATED. 60 PER CEXT LIGHT BROWX, 40 PER CEXT MAPLE.
A. T. Charron, Ottawa, Canada
E. B. HoUand. Amherst. Mass
C. P. Moat, Burhngton, Vt
M. C. Albrech, Washington, D. C
E. Monroe Bailey, New Haven, Conn -J
W. B. Pope, Concord. N. H
A. P. Sy, Buffalo. X. Y
C. H. Jones, Burlington, Vt
A. G. Xiekles, Agricultural College, X.Dak
A. Valin, Ottawa. Canada
J. A. MiUer, Buffalo, X. Y
R. M. West, St. Paul, Minn
F. T. Shutt, Ottawa, Canada
Average
0.59
.47
.58
.47
.47
.68
.67
.59
.56
.57
6.31
.52
.50
.51
.58
.60
0.47
.39
.43
.31
.32
.52
.51
.43
.45
.41
' .23
.38
.36
.39
.45
0.12
, 0.48
.08
.52
.51
.51
.16
.18
.15
.16
.16
.34
.16
.32
.16
.48
.11
.53
.16
.52
.08
a. 37
.14
.42
.14
.44
.12
.41
.13
.23
0.34
.38
.33
.34
.34
.38
.38
.58
.38
.31
a. 34
.34
.34
.48
-40
.11
.11
.04
.04
.13
.05
.12
a. 01
.08
.07
56
.08
17
Table 3.— Kcsulls of coopcratirc maple sugar vork — ('ontiiUKKJ.
No. 3.— PURE MAPLP: SUGAR.
Analyst.
Total
ash.
Soluble
ash.
Insoluble
ash.
Alkalin-
ity of
soluble
ash.
Alkalin-
ity of
insoluble
ash.
Malic
acid
value.
A T Charron Ottawa, Canada
Per cent.
0.71
.73
.72
/ .82
fs
\ .78
.75
.74
.78
.71
f .61
i .62
.86
.70
.74
Per cent.
0.43
.44
.40
.58
.59
.44
.44
.41
.46
.50
.49
.34
.32
.58
.39
Per cent.
0.28
.29
.32
.24
.24
.34
.34
.34
.28
.28
.22
.27
.30
.28
.31
cc.
0.66
.60
.53
.36
.44
.53
.55
.58
.60
.64
a. 48
.42
.44
.64
.41
cc.
0.74
.75
.64
.70
.70
.75
.74
.88
.60
.64
a. 54
.60
.64
.82
.76
0.64
E 13 Holland. Amherst Mass
C. P. Moat, Burlington, Vt
.42
M. C. Albrech, Washington, D. C
E. Moni-oe, Bailey, New Haven, Conn
W B Pope Concord, N. H
.44
.47
.23
.25
.33
A. P. Sy, Buffalo, N. Y
.29
.55
A.G. Nickles, Agricultural College, N.Dak.
A. Valin, Ottawa, Canada
a. 16
.50
.56
.22
J. A. Miller, Buffalo, N. Y
R M West St. Paul Minn
.41
F. T. Shutt, Ottawa, Canada
Average
.73
, .44
.29
.53
.70
.39
a Phenolphthalein used as indicator.
b Not included in average.
Comments by Analysts.
A. T. Charron: The alcohol at my disposal did not contain 95 per cent absolute
alcohol, and I believe my figures on malic acid value are low. Quite a difference in
the appearance of ash was noticed. The ash of No. 3, besides being leafy and fluffy,
contained manganese. All results are the mean of agreeing duplicates.
/. A Miller: Concerning malic acid value, would say that after washing until the
entire filtrate measured 100 cc an additional washing with 75 per cent alcohol showed
the presence of soluble lime salts in the filtrate. Adding 5 or 6 drops ammonia to 100
cc of filtrate, heating one-half hour on bath, and standing overnight, a further precipi-
tate was obtained giving a malic acid value as follows: No. 1, 0.54; No. 2, 0.39; No. 3,
0.43.
A. P. Sy: Characteristic green color absent in above samples. For more accurate
results would suggest that 10 grams be used for incineration.
M. C. Albrech: In addition to the work asked for by the referee the following data
were supplied: Moisture at 70° C. in vacuum. No. 1, 5.12 per cent; No. 2, 6.73 per
cent; No. 3, 7.73 per cent.
A. Valin: To establish uniformity of standards I would recommend that all results
be stated for 100 grams of dry sugar. I also recommend that 35 per cent and 10 per
cent moisture be adopted for sirup and sugar respectively, so as to permit the use of
factors in calculation. The method for lead subacetate precipitate used in the Inland
Revenue Laboratory is as follows: Five grams sugar or sirup are weighed into a test
tube and dissolved in 20 ccof water. Two cc of lead subacetate solution (sp. gr. 1.26)
are added and the solution mixed. After standing a few hours the mixture is filtered
into a sugar tube, washed with warm water four or five times, dried, and weighed. The
precipitate obtained, multiplied by 30.77 for a sirup and 22.22 for a sugar, gives the lead
subacetate precipitate for 100 grams of dry sugar, assuming the sirup to contain 35 per
cent and the sugar 10 per cent of water.
W. B. Pope: The method for malic acid value as outlined gives lower results than
when ammonia is used. Length of time in boiling would also seem to be an important
factor in securing comparable results, as it is evident that, unless the ash is finely
divided, solution on heating is not quite so rapid a process as might possibly be inferred.
Using the method for malic acid value outlined on page 321 of the Eighteenth Annual
31104— No. 105—07 2
18
Report of the Vermont Experiment Station, in which the solution is made slightly-
alkaline with ammonia, the following results were obtained: No. 2, 0.65; No. 3, 0.85.
E. M. Bailey: You will note that one determination of malic acid value in No. 3,
using 0.5 cc of ammonia, gave a result of 0.77. This is much higher than the others,
thus corroborating your experience. The difference between sucrose and total solids
is considerably greater in No. 3 than in the others. We find that in samples of maple
products, consisting largely of cane sugar, this difference is very slight indeed.
Table 4. — Additional data on referee^ s samples determined by Bailey.
Determinations.
Total solids (per cent) .
Sucrose (per cent)
Lead number
Malic acid value, using 0.5 cc ammonia.
Sample 1.
94.12
89.70
0.66
.63
.56
Sample 2.
92.20
86.46
1.37
1.35
1.31
Sample 3.
89.36
81.05
2.00
2.05
2.09
2.13
.77
Comments by Referee.
With but few exceptions the results on the ash work are very satisfactory, and in
all cases give data sufficient for the judging of purity. The figures showing alkalinity
of the soluble and insoluble ash. vary more than is desired. The reason for this, par-
ticularly in the case of soluble ash, is not clear.
The malic acid values again show surprising variations. It is worthy of note, how-
ever, that if the individual results by the different analysts are compared they show,
without exception, the desired distinction between the pure and the adulterated
samples. Taken in connection with the ash figures they show strikingly the sophis-
tication of Nos. 1 and 2.
The conclusions of your referee as to the effect of ammonia on malic acid value are
also confirmed by the additional data supplied by Messrs. Miller, Pope, and Bailey.
(See comments by analysts.)
Attention was called last year to the special method of boiling and filtering and its
use in detecting relatively small amounts of cane sugar added to maple goods. « The
following results were obtained by its use on the association samples:
Table 5. — Results obtained using special method of boiling and filtering .
Determinations.
Total ash (per cent)
Soluble ash (per cent)
Insoluble ash (per cent)
Alkalinity soluble ash (per cent) . .
Alkalinity insoluble ash (per cent)
Malic acid value
C. P. Moat.
0.24 0.41
.14 .27
.10
.20
.24
.02
0.50
.27
.23
.39
.45
.23
C. H. Jones.
0.21
.13
.26
.18
0.44
.30
.14
.39
.31
.07
0.52
.32
.20
.44
.40
.33
Among other methods, useful in detecting the sophistication of maple goods,
which should receive the attention of this association are the following:
(1) The determination of lead number devised by A. L. Winton.& This consists
in determining indirectly the amount of lead in the precipitate formed in maple
products by basic lead acetate. The method is easy of application and not very
lengthy, particularly if sucrose is also to be estimated polariscopically. In this way it
« Eighteenth Annual Report of the Vermont Agricultural Experiment Station, 1905,
p. 328.
b J. Amer. Chem. Soc, 1906, 28: 1204.
19
is thought that the varying results from the two centrifugal methods in common use"
can be stated on a constant gravimetric basis.
(2) Albert P. Sy has devised a method of determining lead directly in the lead
acetate precipitate. & As briefly outlined in a letter to the referee it is as follows:
Dissolve 25 grams of sirup or sugar in 300 cc of water, boil, add 20 cc of neutral lead
acetate, let settle; filter, wash with 200 cc of water at 75° C, digest with aqua regia,
add sulphuric acid, heat to fumes, cool, dilute, add alcohol, settle; filter, ignite, and
weigh as lead sulphate. Calculate to 100 grams of material. The lead sulphate figures
on the association samples by this method were as follows: No. 1, 0.0704; No. 2, 0.0684;
No. 3, 0.288.
(3) Method of Hill and Mosher. c This method consists in removing the lead from
the lead acetate precipitate by hydrogen sulphid, filtering, boiling, and titrating
filtrate with tenth-normal alkali.
Recommendations.
The referee makes no formal recommendations, but suggests that if the work on
maple products is continued that the three methods just mentioned be thoroughly
tried. Also that consideration of the malic acid value be continued until a modifica-
tion giving less variation among different operators is secured.
KEPOET ON FEUIT PKODUOTS.
By Hermann C. Lythgoe, Associate Referee.
The following methods for the examination of lime juice, compiled by the referee,
are in the main very satisfactory:
Specific gravity. — Make the determination of specific gravity in the usual manner.
Acidity. — Titrate 7 grams (6.8 cc if the gravity is about 1.04-1.05) with tenth-normal
sodium hydroxid, using phenolphthalein as indicator. If the sample is colored, dilute
with water. The burette reading divided by 10 gives the per cent of citric acid.
Solids. — Evaporate 5 grams of the sample in a flat-bottomed platinum dish for 2
hours over a boiling water bath, cool, and weigh.
Ash and soluble ash. — Make the determinations of ash and soluble ash on the residue
obtained above as directed in Bulletin 65, p. 55. Examine for colors and antiseptics
as usual — the antiseptics to be suspected are salicylic, benzoic, and sulphurous acids,
and mixtures of these substances.
The determination of the alkalinity of the ash is given by several methods in the
reports of this association, all of which give different results. This is a very valuable
figure in the study of all fruit and vegetable products and there should be more
uniformity in the methods. The method preferred by the referee is as follows:
AlJcalinity of ash. — Evaporate 50 or 100 grams of the sample to dryness and prepare
the soluble ash as described above. Add an excess of tenth-normal sulphuric acid,
boil to expel the carbon dioxid, and titrate back with tenth-normal sodium hydroxid,
using phenolphthalein as the indicator. Express results as cubic centimeters of tenth-
normal acid required to neutralize the soluble ash of 100 grams of sample.
It is recommended that this association take some action to obtain greater uniformity
in the determination of the alkalinity of the ash.
Mr. BiGELOW. The determination of the alkahnity of the ash is
very important in the work on fruit products, and two points espe-
cially should be considered and a uniform procedure established.
The first one is to decide whether hot or cold water shall be used, and
the second and more important point is to determine what indicator
a J. Amer. Chem. Soc, 1904, 26: 1532; also Vermont Exper. Stat. Rpt. No. 17,
1904, p. 454.
& J. Franklin Inst., July, 1906, p. 71.
c Technology Quarterly, June, 1905, 18 (2): 147.
20
shall be employed. At present methyl orange is used for some
products, phenolphthalein for others, and litmus in the case of deter-
mining tartaric acid compounds in wine b}^ the titration of the ash.
Because of this condition we have great difficulty in understanding
published results and comparing the figures. Although Mr. Davidson
as chairman of the committee on unification of terms for reporting
analytical results has tried to get an expression of the opinion of the
members of the association on this subject, only a very few of the
members interested in the examination of foods have replied. As
this committee and the committee on the revision of methods are
preparing reports, it seems ver}^ important that the matter should
receive the attention of the association.
EEPORT 0^ BEER. '
By H. E. Barnard.
The report of Mr. Barnard, Indiana State board of health, on
methods of beer anal3"sis is published as Circular No. 33 of the Bureau
of Chemistry, being too long for incorporation in the Proceedings.
The report includes results obtained in cooperative work according to
various modifications of the provisional methods and a statement of
the methods, based on these results, which are recommended for
adoption as official at the meeting of 1907.
EEPOET ON DISTILLED LIQUORS.
By C. A. Crampton, Associate Referee.
Under date of April 14, 1906, the associate referee sent a circular letter to all analysts
likely to cooperate in the work. FaA^orable replies were received from nineteen
analysts, two them of arriving too late, and seventeen sets of samples were sent out,
accompanied by the following letter of instructions:
May 2, 1906.
Dear Sir: * * * As stated in ray circular letter dated April 14, I expect to
devote the work this year chiefly to the fusel oil and ethereal salts determinations.
The work last year upon the detection of factitious whiskies by methods for arti-
ficial coloring was so satisfactory that it does not seem necessary to go over the same
ground again.
I will ask you, therefore, to determine the fusel oil in each sample, using both the
Roese method and the modified AUen-Marquardt method, adopted provisionally at
the last meeting of the association. You will also please use the modified method
for esters adopted at the same meeting.
I inclose herewith a copy of Circular iSio. 26, Bureau of Chemistry, Department of
Agi'iculture, which will give you the newly adopted methods.
Should you have time and material left after making the above determinations, I
should be very glad of any further results on the samples, and will leave it to your
option to make tests for coloring matter or the routine determinations of alcohol,
extract, ash, etc. I would suggest that the determinations of aldehydes and furfural
can be made in the distillate which is used for ethereal salts. * * *
Respectfully,
C. A. Crampton,
Associate Referee on Distilled Liquors.
Reports have been received upon twelve sets of samples, and the results are included
in the tabulation which follows. As indicated in the circular letter, three samples
were sent to each analyst, numbered 1, 2, and 3.
No. 1. Straight Bourbon whisky.
21
No. 2. Artificial whisky, made by adding coloring matter to alcohol, with an addi-
tion of 0.3 per cent by volume of amyl alcohol, or 0.243 per cent by weight.
No. 3. Straight rye whisky.
The following table gives the results obtained by different analysts and the varia-
tions and averages. All results are calculated to grams per 100 cc, using 0.810 as the
specific gravity of amyl alcohol (fusel oil).
Results of cooperative work on whishy samples.
SAMPLE NO. 1.— BOURBON WHISKY.
Fusel oil.
Esters.
Aldehydes.
Analyst.
Roese
method.
Provisional
method.
Furfural.
G. E. Boiling
0.389
0.170
.178
.226
.191
.167
.215
.309
.133
.064
.185
0.072
.062
.083
E. M. Chace...:
0.032
0.001
Julius Hortvet
.155
.243
.235
.069
.069
.070
.072
.084
.040
.005
L. M. Law
H. M. Loomis
0. S. Marckworth. . .
.020
.006
W. B. Pope
R. E. Stallings
H. A. Weber. .
.196
.405
Present.
.001
T. D. Wetterstroem
.132
.079
SAMPLE NO. 2.— ARTIFICIAL WHISKY CONTAINING 0.243 PER CENT AMYL ALCOHOL
BY WEIGHT.
G. E. Boiling
E M. Chace
0.389
0.170
.160
.220
.153
.181
.293
0.013
.002
.033
Trace.
None.
Edward Guderaan
.153
.277
.277
A Lasche
.005
.004
.008
.007
.014
.0
0.007
L M. Law
0 S Marckworth
.002
None.
W. B. Pope
.138
.225
R. E. Stallings
H. A. Weber
.298
.381
None.
0.0
T. D. Wetterstroem l
.177
.020
SAMPLE NO. 3.— RYE WHISKY.
G E Boiling
0.462
0.200
.202
.248
.245
.208
.259
.268
.179
.211
.255
0.073
.063
.077
E. M. Chace
0.033
0.003
Edward Gudeman
Julius Hortvet
.266
.351
.214
.074
.079
.067
.085
.072
.044
.005
L M. Law
H M Loomis
0. S. Marckworth
.018
.013
W. B. Pope
R E Stallings
.260
.389
Present.
.004
H. A. Weber
T. D. Wetterstroem
.176
.093
AVERAGES AND VARIATIONS.
Sample
No.
Fusel oil.
Data.
Roese
method.
Provi-
sional
method.
Esters.
Alde-
hydes.
Furfural.
Average
f 1
J 2
0.270
.296
.324
.135
.093
.138
.115
.143
.110
0.179
.191
.223
.130
.102
.045
.047
.053
.047
0.070
.010
.073
.013
.023
.012
.030
.010
.029
0.019
.002
.019
.013
.005
.014
.014
.002
.014
0.003
000
Highest variation
3
1
2
3
1 ^
1 3
.007
.003
.006
.002
.004
22
COMMEXTS BY AXALYSTS.
George E. Boiling: Regarding the Allen-Marquardt method there appears to have
been a wholesale loss durmg some part or parts of the process. Would mercm'v seals
be permissible dm-ing oxidation, etc.?
Edward Gudeman: The method for fusel oil was only slightly varied, using 300 cc
instead of 100 cc of whist\'. Everj-thing else was used in same proportion. The
carbon tetrachlorid after washing with the salt solution was made up to 300 cc. and
of this 100 cc were used for oxidation and distillation. This was done so as to have
identical samples for testing whether the use of tinfoil was necessary and actually
took part in the reaction. My results show that the tinfoil does not take part in the
reaction and simply acts as protection for the corks.
A. LascJie: Submits a paper giving the results obtained on a number of samples
containing known quantities of various fusel-oil constituents by the two methods.
This paper is published in full in Lasche's Magazine. September. 1906. page 105.
Following are his conclusions :
The conclusion to be drawn from these investigations, it is apparent, is that the
results obtained according to the AUen-Marquardt method do not represent the
quantities of higher alcohols or fusel oil contained in distilled liquors, whereas the
Roese method results are very reliable and practically correct.
I do not stand alone m my opinion as to the relative merits of these two methods,
and see fit to quote here from personal communications received from such distin-
guished investigators as Prof. Dr. J. Koenig, Mlinster, Germany, and Prof. Dr. Karl
Windisch, Hohenheini. Germany. tJ
I am satisfied that our restilts are conclusive as to the application of the Allen-
Marquardt method for fusel-oil determination in whisky. The Roese method is
undoubtedly the most practical and reliable method known to-day and should be
retained as the ofiicial method for fusel-oil determinations.
L. 31. Law: From the results obtained thus far in a study of ftisel oil. the provisiona
method seems the more promising for obtaining concordant and uniform results. The
defect of its giving figures somewhat low can possibly be oA*ercome by finding a factor
for the oxidation process which, as is well known, is not a complete one. It is, at any
rate, the method for the average analyst with the average laboratory facilities.
H. M. Loomis: The liquors were saponified in all cases by standing cold overnight.
Some difficultj^ was experienced due to foaming in the first distillation, which tannin
did not obviate; also found trouble in getting a sharp end point in making the valeric
acid solution neutral to methyl orange.
To find whether any loss of valeric acid took place during the eight hours heating in
the apparatus used, an experiment was made which showed that no such loss took
place during the oxidation process.
0. S. Marckworth: I find the fusel oil method quite tedious, but prefer it to the
Roese, providing I can duplicate the results, which I will attempt in the next two
weeks. The end point with methyl orange is quite difficult to determine.
E. E. Stallings: The results by the two methods did not agree A^erj' closely, as will
be seen from the table. It seems that the pro~\T.sional method (modified Allen-
Marquardt) is quite a long and tedious one and there are many chances for error.
a Doctor Koenig says : '' Die praktische und auch gleichzeitig ztiveraliissigste Methode
ztu" Bestimmung von Fuselol bezw. Amyl-Alkohol in alkoholischen Getriinken ist
das Yerfahren von Roese. Das Verfahren von AUen-Marquardt wird wegen seiner
Umstandlichkeit kaum mehr gebraucht und hat auch angehlich inanche 3f angel."
"Windisch says: "'Ich habe mit dem Verfahren von Roese die besten Ergebnisse
bekommen und bin auch Heute noch der Meinung. dass dieses Yerfahren besser ist
als die tibrigen Yerfahren, und auch das Verfahren von AUen-Marquardt.''
23
Comments by Associate Referee.
The work was limited this year to a study of the fusel-oil determinations, with the
hope of arriving at definite conclusions concerning the reliability of the official methods,
but the results are meager and disappointing. Only five of the twelve analysts
reporting used both methods.
The figures obtained on the same sample by the different methods vary considerably,
but not more so than the results obtained by different analysts with the same method.
Results by the Roese method seem to run high, every analyst except one obtaining
results higher than the known quantity in sample No. 2. Results by the Allen-
Marquardt method, on the other hand, run low, all the results being below the known
quantity in No. 2, except in one case. There seems to be little doubt that the oxida-
tion process in the latter method is not complete. A further study seems to be
essential.
The results for esters agree well, except in the artificial whisky. No. 2. From the
fact that the process is simply one of distillation and titration in a perfectly clear
solution, it would seem that better agreement should have been reached. The method
is undoubtedly superior to the previous method, which involved a double titration in
a colored solution.
Recommendations.
I have no recommendation to make as to changes in methods. For the ensuing year
I would recommend that a further study be made of the determination of fusel oil.
Literature on Fusel Oil.
Results of Fusel Oil Determinations, according to the Allen-Marquardt method as
modified by Schidrowitz: Lasche's Magazine for the Practical Distiller, Vol. IV, No. 4,
page 105.
The Roese-Herzfeld and Sulphuric Acid Methods for the Determination of the
Higher Alcohols — A Criticism: V. H. Veley in Journal of Society of Chemical Industry,
Vol. XXV, No. 9, page 398.
No report was received from the referee on vinegar, but the follow-
ing paper on the subject was presented:
rULLEE'S EAETH TEST TOE OAEAMEL IN VINEGAE.
By W. L. Dubois.
The fuller's earth test for caramel appears in a number of publications covering
methods for food analysis, and has been used quite generally for the detection of added
caramel in cider vinegar. In some cases the method has been published with no state-
ment of the precautions necessary in its manipulation, nor the limitations to which
it is subject. In order to investigate these points, and if possible prescribe conditions
under which it could be applied with certainty, the work described in this article was
undertaken.
Fifty samples of pure cider vinegar were obtained from farmers in Pennsylvania
through the State Dairy and Food Commission. Of these, eleven were selected, dif-
fering as much as possible in physical appearance. Five vinegars made by the author
in 1905 were also included in this experiment.
Samples of fuller's earth were procured from several supply houses and from a num-
ber of food chemists, the purpose for which the samples were desired being stated.
The method was applied as follows:
Fifty cubic centimeters of vinegar and 25 grams of fuller's earth were measured
into a 250 cc beaker, stirred thoroughly and allowed to stand one-half hour. The
24
mixture was then filtered through a dry folded filter, and the color of the filtrate com-
pared with that of the untreated vinegar, filtered in the same way. Color comparisons
were made in a Duboscq colorimeter. In the talble below the results are expressed
as per cent of the total color removed by fuller's earth. The last five vinegars in the
table were made by the writer.
Amount of color removed by 9 samjples of fuller's earth from vinegars.
Vinegar,
Nos.
Fuller's
earth numbers.
15663.
15664.
17038.
17039.
17040.
17042.
17043.
17045.
17080.
17046
17047
17048
17049
17050
17055
17056
17057
17058
17059
17060
17512
17513
17514
17515
17516
Per cent.
29.6
41.9
34.0
Per cent.
38.0
46.8
20.0
31.6
36.7
0.0
68.0
28.0
40.0
44.0
G3.1
Per cent.
24.0
48.0
30.0
0.0
0.0
7. 7
67.9
27.2
25.0
20.0
65.0
Per cent.
44.8
60.0
44.0
20.0
4.0
52.0
70.0
32.0
25.0
30.0
66.6
57.9
43.4
41.2
37.5
41.2
Per cent.
28.0
42.0
22.2
0.0
0.0
20.0
66.0
12.0
24.0
4.0
63.1
42.9
24.7
13.0
0.0
37.5
Per cent.
31.2
42.0
25.0
20.0
23.3
36.8
60.7
20.0,
36.0
10.0
53.3
0.0
16.7
0.0
Per cent.
18.4
43.2
25.0
0.0
0.0
33.3
66.6
0.0
4.0
25.0
63.1
35.9
21.8
0.0
11.0
32.5
Per cent.
50.0
50.0
4.2
0.0
Per cent.
0.0
48.3
20.0
0.0
70.8
40.5
66.6
30.0
20.0
23.1
67.6
63.4
66.6
50.0
71.4
45.5
71.9
56.2
38.9
57.6
1
Discussion op Results.
The color removed from pure cider vinegar by fuller's earth varies, according to the
figures in the table, from none to 72 per cent. No one sample of fuller's earth can be
selected from the above as giving uniform results. For instance, Nos. 15663 and 15664
remove no color from vinegar No. 17055, thereby indicating that it is pure, while from
vinegar No. 17056, which is just as pure as No. 17055, 71 per cent and 68 per cent,
respectively, of the color are removed.
Vinegar No. 17050 would be indicated as pure by treatment with earths Nos. 17038,
17040, and 17043, while its quality would be doubtful according to earth No. 15663
and condemned by No. 15664. Again, earths Nos. 17038, 17040, 17043, which give
uniform results on vinegar No. 17050 and comparable results on several others, are at
wide variance on vinegar 17057. Earth No. 17080 is one received from a chemist who
uses this test and declares it reliable if 25 to 30 per cent be allowed for the color which
fuller's earth will remove from pure vinegar. As shown in the table, it removes from
one sample of pure cider vinegar no color at all, while taking out as much as 72 per
cent of color from another sample. There seems to be no uniformity in the data, and
it is impossible to select any one of the fuller's earths tried which could be relied upon
to give truthful results. In the writer's estimation the method is unreliable, and
should only be used as a preliminary test. If no color, or only a small percentage of
color, be removed, the analyst is reasonably safe in pronouncing the sample pure. On
the other hand, if all the color disappears he is equally secure in declaring caramel
present. But for the large number of vinegars which lose from 25 to 75 per cent of
their color when treated with fuller's earth the data obtained by this test are not final,
and it is necessary to subject the vinegar to further treatment before a conclusion
regarding the presence of caramel can be reached.
Mr. Crampton. The fuller's earth test for caramel in vinegar, as
published by Mr. Simons and myself some time ago, was put forth
only in a tentative way, as it had been used mainly upon distilled
spirits and not applied to many samples of vinegar. This paper is
the first report I have heard of any extensive work being done with
25
the method as appHed to vinegars, and I think undoubtedly the test
should be considered only as corroborative and be accompanied by
other tests.
EEPOET ON TLAVOEINa EXTEAOTS.
By E. M. Chace, Associate Referee.
The work for the past year has been confined to testing the proposed methods on
the determination of the aldehydes (citral) in lemon oils and extracts. The following
methods were given a preliminary trial by the referee: Sadtler's,^ Romeo's & modifi-
cation of Sadtler's, and Berte's.c Of these, Sadtler's gave promise of being the best.
The method, in brief, is as follows:
Five to 10 grams of the oil are weighed into an Erlenmeyer flask and, after neutraliz-
ing, 25 or 50 cc of a 20 per cent solution of neutral sodium sulphite are added. Rosolic
acid or phenolphthalein may be used as indicator. On mixing the oil and sulphite
solution a red color immediately forms, owing to the alkali set free by the reaction.
This is neutralized with half-normal hydrochloric acid from time to time, the flask
containing the mixture being surrounded by boiling water. The reaction is complete
in about half an hour. The citral is found from the amount of acid used to neutralize
the alkali set free according to the following reaction:
C9Hi5CHO+2H20+2Na2S03=C9Hi5CHO(NaHS03)2+2NaOH.
The method has two serious defects: (1) The alkali formed must be neutralized
immediately, or at the temperature of boiling water it attacks the aldehyde sulphite
compound regenerating citral; and (2) the end point is obscure on account of dissocia-
tion of the sulphite in solution. On this account the final neutralizing must be done
in the cold. With the unmodified method it was impossible to obtain concordant
results.
It was found that the delicacy of the end point could be considerably increased by
the addition of salt to the solution, but even then the results seemed uniformly low,
although the greatest care was used in neutralizing the alkali as soon as formed.
The following results were obtained on solutions of citral in alcohol :
Determinations of citral in alcohol solutions {Sadtler's method).
Amount
Amount
Percentage
added.
recovered.
recovered.
Grams.
Grams.
Per cent.
0.500
0.366
73.2
.215
.206
96.0
.510
.479
92.0
.510
.444
87.0
.510
.401
80.0
Samples made up in limonene, upon which the citral had been previously deter-
mined by the same method, gave correspondingly low results.
The modification of this method by Romeo consists in using a solution of potassium
acid sulphite (the acidity of which equals the alkalinity of half-normal alkali) to neu-
tralize the alkali formed. After neutralizing the sodium sulphite, a definite volume —
about 20 cc — of the potassium acid sulphite is first added and then the neutralized
lemon oil. This mixture is heated on the water bath under a reflux cooler, with
frequent shaking for three hours, is then removed, cooled, and the excess of acidity
due to the unneutralized potassium acid sulphite titrated with half-normal sodium
a J. Franklin Institute, December, 1903; February, 1904.
bChem. and Drug., 1905, 408.
c Chem. Ztg. 29, 805, Abs. in Analyst, October, 1905.
26
hydroxid. In this \ray the method overcomes the error due to the presence of free
alkali, but it stUl retains the difficulty of an obscure end point.
Fair results Tvere obtained when solutions of pure citral were used and s^:>dium
chlorid added before titrating.
Results using chemically pure citral and adding sodium cldorid.
Amount
^ added.
Amount
found.
Percentage
found.
Grams.
0.540
.So
.510
..525
..5.SS
.4S6
Grams.
0.531
.549
.534
.554
.527
.496
Per cent.
98
105
105
The method failed completely, however, when the citral was added to limonene
solutions. Thus a mixtiu'e of limonene and citral, theoretically containing 6.62 per
cent of citral. gave 5.72 per cent. 7.02 per cent. 5.23 per cent, and 5. 86 per cent.
Berte's method depends upon the polarization of the oil before and after the removal
of the citral by potassium acid sulphite. Ten cc of the oU are emulsified with 50 cc
of potassium acid sulphite and heated for ten minutes on the water bath, cooled, re-
emulsified, heated for five minutes, cooled, and the oil separated by means of a separa-
tory funnel. It is then thoroughly washed and dried by shaking with anhydrotts
sodium sulphate. The original oil and this citral-free oil are now polarized at the
Kime temperature and the citral calculated by the following formula :
^ 100 <B-A)
"~ B
in which A is polarization of the original oil. B the pc-larization of the citral-free oil,
and C the per cent of citral. Only negative results were obtained by this method;
the citral apparently is not removed by the treatment.
Xo methods depending upon the use of the cassia flasks were tried, as the consensus
of opinion seems to be that those methods do not give accurate results on lemon oil.
owing either to the resins present or to the fact that the citronella aldehyde present
forms a sulphite compound which is insoluble in water.
A method depending upon the reduction of ammonia cal silver nitrate was also tried,
the results being uniform where citral alone was used, but very irregular on mixtures
of limonene and citral.
The method finally adopted for collaborative work was a modification of that of
Medicus.a and the following instructions & and samples were sent to the eight chemists
who had expressed a desire to collaborate.
A was a commercial lemon oU.
B was a solution of c. p. citral in impurified alcohol containing 0.25 per cent by
weight.
"C was a solution of limonene containing 0.12 per cent of citral.
D was an extract made by adding 5 grams of A to 100 grams of alcohol.
The instructions sent with the samples were as follows:
a Forschungs-Berichte liber Lebensmittel. 1895, 2 : 299.
& See also J. Amer. Chem, Soc 1906. -28: 1472.
27
COLORIMETRIC METHOD FOR THE DETERMINATION OF CITRAL IN LEMON EXTRACTS.
Reagents.
Standard citral solution. — Made by dissolving 1 gram of c, p. citral in 1 liter of 50
per cent alcohol.
Fuchsin solution. — Dissolve one-half gram of fuchsin in 100 cc of water and add a
sulphurous acid solution containing 16 grams of SOg. Allow to stand until the color
disappears and make up to 1 liter. This solution deteriorates on long standing.
Aldehyde-free, 95 per cent alcohol. — Various methods may be used for purifying the
alcohol. The following method gave the best results in this laboratory:
The alcohol is allowed to stand over caustic alkali for a day or two and is then
distilled. This distillate is boiled for three hours under a reflux cooler with 25 grams
per liter of meta-phenyl-diamine-hydrochlorate and finally distilled. The purified
alcohol gives only a trace of color when treated with the fuchsin solution at 15° C.
Note. — All reagents must be cooled to about 15° C, and used at that temperature.
While it is not necessary that they be kept at exactly 15°, they must all be kept at
the same temperature. A cooling bath such as is described by Given « is used in this
laboratory. This bath gives abundant room for all reagents and comparison tubes.
Manipulation.
Approximately 2 grams of lemon oil are weighed in a small flask, transferred with
alcohol to a 100 cc flask, and made up to the mark with alcohol at the temperature of
the cooling bath. (For lemon extracts, weigh 20 grams and make up to 50 cc.)
It is well to make up a second solution after calculating the first, so that this solution
shall have the same strength as the standard, i. e., 1 mg per cubic centimeter, and to
do the final work on this.
In making up the standards and samples for comparison, add first the lemon prod-
uct or citral, next 20 cc of aldehyde-free alcohol, then 20 cc of fuchsin solution, and
finally make up to the 50 cc mark with alcohol. Allow to stand in the cooling bath
for ten minutes, and read.
Four milligrams are preferred as a standard to use with colorimeters, and 2 mg
where comparisons are made in Nessler tubes. Final comparisons should be made on
solutions of approximately the same depth of color, as the increase in color is not exactly
proportional to the amount of citral present. Calculate the results as per cent by
weight.
The samples sent are as follows: A, lemon oil; B, citral solution; C and D, lemon
extracts. The method has given very satisfactory results on lemon extracts; on lemon
oils it is but fair, owing to the great multiplication of the error. Working with alcohol
of such high percentage it is very evident that temperature is a factor which must
be closely watched. In this laboratory we cool even the pipettes before using them.
The time given for developing the color (ten minutes) is optional; fifteen minutes may
be used. The essential requirement is that the standards and samples should stand
the same length of time at the same temperature.
But four reports were received, one of the collaborators having worked under con-
ditions so widely different from those prescribed by the method that his results have
not been tabulated.
"J. Amer. Chem. Soc, 1905, ;^7 (12): 1519.
28
Cooperative ivorh on the determination of citral in lemon extracts {colorimetric method).
Analyst.
Samples.
B.
C.
Geo. E. Boiling, sewer, health, andjwater department, Brock-
ton, Mass
A. F. Seeker, food inspection laboratory, port of New York. .
H. J. Holland, food inspection laboratory, San Francisco, Cal .
E. M. Chace, associate ^referee, Bureau of Chemistry, Wash-
ington, D. C ." .,.
6.40
4.53
5.00
(a)
5.29
5.02
5.16
Per cent.
0.80
[ .275
.312
.284
.279
.282
Per cent.
0.18
.115
.147
.134
.134
.121
Per cent.
0.28
.23
/ .217
I .258
f .262
.267
.280
.264
I .277
a Not reported.
It is recomin ended that the collaborative work on the method be continued for
another year.
EEPOET ON BAKING POWDEES.
By W. M. Allen, Associate Referee.
As no recommendations regarding the work on baking powders and baking chemicals
for this year were made in 1905, it was the intention of the associate referee to take up
the recommendations made in 1904, but owing to other duties this work was reached
too late for cooperative investigation. The referee had a large number of baking
powders on which the available carbon dioxid was to be determined, the total and
residual not being required except to give the available by difference, and as these
determinations were to be made in duplicate it was deemed advisable to devise some
method for determining the available carbon dioxid directly, without making the
^total and residual determinations. For this purpose the following method, based on
the method of McGill and Catlin for the determination of the residual carbon dioxid
as described by Winton in the Provisional Methods for the Analysis of Foods,"
was tried:
DETERMINATION OF AVAILABLE CARBON DIOXID WHEN THE TOTAL AND RESIDUAL
AMOUNTS ARE NOT REQUIRED.
Weigh 2 grams of the baking powder into a dry flask of about 150 cc capacity, and attach
the flask to a Heidenhain's apparatus for the determination of carbon dioxid. Through
the funnel tube, which must have a stop cock, run into the flask containing the sample
about 30 to 40 cc of cold water. The condenser of the apparatus, to which the flask
is attached, should have some little play so that the flask may be shaken with a slight
rotary motion until the sample is thoroughly mixed with the water. Place the flask
in a small water bath, filled with cold water, by raising the bath up under the flask
until the latter is about half submerged. The water bath should be just large enough
to conveniently accommodate the flask. Heat the bath to boiling, taking about eight
to ten minutes to reach that temperature, so that the heat in the flask will be gradual.
Continue the boiling until the sulphuric acid in the indicator of the apparatus begins
to recede, showing that the operation is complete. Then attach and start the aspira-
tor and open the funnel tube, which should be guarded by a soda-lime absorption tube.
Complete the determination as directed under total carbon dioxid, Provisional
Methods, Bulletin No. 65, above mentioned, page 98, using the precautions there noted.
By this method results agreeing closely with those given by the provisional methods
are obtained. If the amounts of residual and total carbon dioxid are required, the
determinations are completed as follows:
a U. S. Dept. Agr., Bureau of Chemistry, Bui. 65, p. 104.
29
RESIDUAL CARBON DIOXID.
After the available is determined and the soda-lime absorption tubes are replaced
in the apparatus, run into the flask containing the sample about 40 cc of hydrochloric
acid (sp. gr. 1.15) to set free the residual carbon dioxid. Boil the acid very gently
for a few minutes, and complete the determination as above directed under available
carbon dioxid.
TOTAL CARBON DIOXID.
To obtain the amount of total carbon dioxid, add the residual to the available.
The analytical results obtained in the trial of this method are given in the following
table :
Comparison of the determination of available carbon dioxid in baking powders by the direct
and the provisional methods.
Available carbon
dioxid.
Sample.
Available carbon
dioxid.
Sample.
Provisional
method,
by differ-
ence.
Direct
method.
Provisional
method,
by differ-
ence.
Direct
method.
Cream of tartar baking
powder, starch filler '
Phosphate baking powder,
starch filler ]
Per cent.
12.29
11.37
11.40
12.21
9.54
13.54
9.74
10.26
Per cent.
12.20
11.40
11.36
12.18
9.60
13.58
9.70
10.18
Alum baki ng powder,
starch filler 1
Per cent.
15.90
10.99
8.01
10.99
11.22
10.55
8.01
Per cent.
15.78
10.92
8.07
Alum-phosphate baking
powder, starch filler "
10.92
11.24
10.54
8.03
EEPORT ON FATS AND OILS.
(Cooperative work on the cloud and the cold tests for 1905 to 1906.)
By L. M. ToLMAN, Associate Referee.
This work was begun two years ago by sending out a preliminary circular giving the
various methods used in different laboratories, asking for suggestions. It was found
that three general types of methods were in use for making the so-called cold test:
(1) The cold test as practiced for cotton-seed oil used for salad purposes, in which
the oil is placed in a bottle and allowed to stand at a definite temperature for a
definite time, when there must be no formation of crystal and the oil must be per-
fectly clear.
(2) The cloud test as practiced in the Armour laboratories, in which the tempera-
ture at which the first cloud is formed in the oil as it is cooled is determined.
(3) The flowing test, in which the temperature at which an oil which has been
frozen will flow under definite conditions is determined.
These tests are evidently quite different, they have different objects in view, and
should have different names, and it has been suggested that the following nomencla-
ture be used: The first test to be called the cold test; the second, the cloud test; and the
third the flowing test. These names are suggestive, and will be adopted in the future
in this work unless some reason is suggested for a change.
In a second circular sent out this year a summary of the suggestions that had been
made was given, and the following methods were offered for trial, four samples of oil
being sent to each of the collaborators:
30
CLOUD TEST.
The cloud test is given by Mr. Manns as follows :
(1) The oil must be perfectly dry, because the presence of moisture will produce a
tui'bidity before the clouding point is reached.
(2) The oil must be heated to 150° C. over a free flame, immediately before making
the test.
(3) There must not be too much discrepancy between the temperature of the bath
and the clouding point of the oil. An oil that will cloud at the temperature of hydrant
water should be tested in a bath of that temperature. An oil that will cloud in a mix-
tm-e of ice and water should be tested in such a bath. An oil that will not cloud in a
bath of ice and water must be tested in a bath of salt, ice, and water.
The test is conducted as follows: The oil is heated in a porcelain casserole over a free
flame to 150° C, stirring with the thermometer. As soon as it can be done with safety,
the oil is transferred to a 4-ounce oil bottle, which must be perfectly dry. One and one-
half ounces of the oil are sufficient for the test. A dry Fahrenheit thermometer is placed
in the oil, and the bottle is then cooled by immersion in a suitable bath. The oil is con-
stantly stirred with the thermometer, taking care not to remove the thermometer from
the oil at any time during the test, so as to avoid stirring air bubbles into the oil. The
bottle is frequently removed from the bath for a few moments. The oil must not be
allowed to chill on the sides and bottom of the bottle. This is effected by constant and
vigorous stirring with the thermometer. As soon as the first permanent cloud shows in
the body of the oil, the temperature at which this cloud occui's is noted.
With care, results concordant to within 1° F. can be obtained by this method. The
Fahrenheit thermometer is used merely because it has become customary to report
results in degrees Fahrenheit.
The oil must be tested within a short time after heating to 150° C, and a re-test must
always be preceded by reheating to that temperature. The cloud point should be
approached as quickly as possible, yet not so fast that the oil is frozen on the sides or
bottom of the bottle before the cloud test is reached.
COLD TEST (MILLWOOD).
Warm the oil until all the stearin is dissolved and filter, through several thicknesses
of filter paper, into a dry 4-ounce Avide-mouth bottle, 1^ ounces of the oil to be tested;
place in a freezing mixture and stir until the oil becomes solid, then cork and leave for
one hour in the freezing mixture. Take the bottle from the freezing mixture, wipe it
dry, and place in a holder of ordinary magnesia, asbestos pipe covering, or any suitable
holder which will insulate the sides of the bottle. The frozen oil is broken up and well
stirred with the special cold-test thermometer previously described, and at every
degree rise in. the temperature the bottle is inverted; continue till the oil will run to
the other end of the bottle. The temperatiu"e registered at this stage is to be considered
the cold test.
The following list of questions was submitted :
For lubricating oils:
(1) Method to be used:
(a) A flowing test?
(6) A clouding test?
(2) Preparation of oil for analysis:
(a) Shall it be dried, and how?
(6) Shall it be filtered?
(3) Method of cooling:
, (a) Shall it be stirred until solid?
(6) Shall it stand a definite time; and if so, how long?
(4) Method of melting:
(a) Shall it be allowed to warm at room temperature?
(6) Shall it be warmed in a bath?
31
As regards salad oils:
Can the cloud test be used for the testing of salad oils, such as winter cotton-seed
oil?
The samples submitted for the work were prime lard oil, neatsfoot oil, grease oil,
and tallow oil. The collaborators were requested to test them as follows:
(1) By the method in use in the respective laboratories.
(2) By the cloud test as given by Mr. Manns.
(3) By the cold test of the Pennsylvania Railroad, as modified by Millwood.
(4) In regard to the other points at issue as far as is practicable.
The need of a special thermometer which can be read without removing from the
bottle was noted by Robert Job, of the Philadelphia and Reading Railroad, and by
J. P. Millwood, of the Brooklyn Navy-Yard. The latter thus describes the ther-
mometer used by him: "The special thermometers used are graduated in degrees
from 0° to 100° F. and are 18 inches long, with the zero point about 7 inches above the
bulb, which brings it outside the bottle."
The following table gives the results reported by the twelve men who sent in their
results :
Results on cloud and cold tests, 1905-6.
Cloud test.
Flowing test (Millwood) .
Analyst.
1.
Prime
lard oil.
2
Neafs-
foot oil.
3.
Grease
oil.
4.
Tallow
oil.
r.
Prime
lard oil.
2.
Neat's-
foot oil.
3.
Grease
oil.
4.
Tallow
oil.
A. V.H.Mory
24 -26
30
13 -15
°F.
51 -53
°F.
72-74
"F.
46
28
"F.
30
5
8
33.5
30 -32
28. 4-32
31
33
31
°F.
45
46
40
48
52 -54
40 -42
43
° .F
75
56
Max H Wickhorst
45
83
W. E. Tinney
26 -27
25 -27
29. 5-33. 5
27 -28
30
24 -26
26. ^28. 4
29
30.2
29 -29.5
29 -30
15 -17
15 -17
17. 6-18
15.5
18
14 -16
15.8
23
23
18
15 -16
50 -52
49 -51
61. 8-63. 5
51
73 -75
72 -74
76 -78
74
45.5
44 -46
43. 5-45
46.5
.46
44
46.4
78
70 -72
R. D. Oilar
82. 4-83. 4
W. D. Richardson...
Wilson Low
84
C. F. Hagedorn
J. E Weber
51 -53
58 -59
48
60
57. 5^58. 5
55 -56
73-75
76
45
57.2
41
44.6
48. 5-49. 5
45 -46
75
80.6
J. P.Millwood
A. H. Schmidt
A.Lowenstein.|^2)"'
76 1 44
76 . 42
77 45 -46
75-76 |46 -47
36
30.2
30. 5-31. 5
29 -30
75
82.4
79 -80
83 -84
Maximum
Minimum
33.5
24.0
23
13
63.8
49.0
78.0
72.0
46.5
28.0
36.0
5.0
57.2
40.0
83.4
56.0
Difference
11.5
10
14.8
6.0
18.5
31.0
17
27.4
FLOWING TEST.
A. H. Schmidt"
A. Lowenstein b
19. 4 9. 2
-46 32 -33 53
-46 32.5-33.5 53
37.4
-54
-54
69
a Congeals sample in a test tube. ' & Sample frozen over night.
The following are the comments and answers given by the various analysts:
Comments of Analysts.
A. G. MANNS.
Cold test. — In the case of prime lard oil, no difficulty is encountered in obtaining
concordant results on the cold test. Different portions of the same oil were tested
under different conditions and the results agreed to within one degree. One portion
was frozen hard and the flowing point taken; another portion was kept in a cooler
having a temperature of — 100° F. for three days, and no difference in the cold test
was noted.
32
YelloTT grease oil gives a -^ide variation of results under different conditions of
testing. One portion of a sample tested in the ordinary way gave a cold test of 39°-41°
F. at a room temperatm-e of 78° F. Another portion frozen tinder the same conditions
along with the fio-st sample gave a test of 46°— 18° F. at a room temperatiu'e of 56°. Still
another portion of the same sample, after having remained for two days in a freezer at
a temperatui-e of — 10° F. gave a cold test of 49°-5r F. Thns a variation of 10° F. was
obtained on the same sample. The result is greatly influenced by the length of time
and temperatm'e of freezing, and the temperature of the room in which the test is
made.
Tallow oil shows the same variation under different conditions, and in the case of
this oil a room ha-\ing a very high temperatm-e must be used in order to obtain the
flowing point, or artificial heat of some kind must be employed.
Concordant results were obtained in testing -exti-a Xo. 1 lard oil. when precautions
were taken to use the same room temperatm-e and the same conditions of freezing.
These restilts show that a different set of specifications must be used for each variety
of oU to be tested, and these specifications must be made very rigid.
Cloud test. — The cloud test gives perfect satisfaction in the testing of all oils met
with in the ever^^day practice of the packing-house and refinery busiaess.
In this method a perfectly defined startirig point is specified, a pei-fectly defined
cooling medium is given, and a definite end point is given. Xo difficulty is encoun-
tered by different observers in obtaining a close agreement on the cloud point, and
the method has the advantage of speed and convenience, so necessai-y where a large
number of determinations must be made daily.
There seems to be no direct relationship between the cloud test and the cold test.
The cloud test is based on the solubility of the stearic constituents in the accompa-
nying palmitic and oleic constituents. The cold test is founded on the viscosity of
the combined constituents. The relationship between these two properties expressed
in degrees of temperatm-e is not constant, because of the varying proportion of the
three constituents in different grades of oil.
The MOlwood test, while it gives exact conditions for fi-eezing the oil. makes no
proA-ision for the temperatm-e of the room in which the test is can-ied out.
As om- experiments show, this factor enters largely into the result, and while it does
not affect the test in the case of prime lard oil it varies the results gi'eatly in the case
of several other oils.
We would answer the questions on page 30 in the following manner:
(1) The cold test would be the more deshable it rigid specifications could be drawn
up to cover the questions of length of time and temperature of freezing and of the room
temperature. We believe the test to be impracticable, however, because the specifi-
cations would have to take into account oils flowing at such a low temperatm-e that it
would requhe a verj-^ expensive and unusual equipment to have rooms in which the test
could be caiTied on. There might be two specifications: one for oils flowing at a high
temperature, and one for oils flowing at a low temperatm-e. That would make the
former a practicable test. The method, too. must always be a lengthy and inconA'en-
ient one, especially as a factory method.
(2) (a) It should be heated rapidly over a free flame to 150° C. (6) It should be fil-
tered if not clear when hot.
(3) (a) It should be frozen rapidly without stuTing. and then be allowed to stand in
the freezing bath for the requisite length of time. (6) The oil must be frozen ver^-
hard throughout; after it is once in this condition, which will require several hom-s at
least, it makes no difference in the result how much longer it is fi-ozen. The length of
time of cooling would var^- greatly with the gi'ade of oil used.
(4) (a) It shoifld be allowed to wai-m up gi-adually at a specified room temperature.
(6) A wai-ming bath would heat the frozen oil superficially thi'ough the sides of the
33
]:>oltle, and allow the whole mass to flow to the other end of the bottle at a temperature
much below the flowing point of the whole mass.
In answer to your question about the test for salad oils, we think the cloud test
should be used.
A. V. H. MORY.
Mr. Mory also notes that the flowing test gives very different results when the condi-
tions of analysis are varied; and that the oil should be dried and filtered. He objects
to the stirring of the oil in the flowing test while the oil is being cooled, as air bubbles
are introduced, which affect the results. He thinks that it would be very diflicult to
so regulate the conditions of the flowing test that it would be satisfactory, and says
regarding the use of the cloud test for salad oil: "I think the cloud test is the proper
one to be used. It should be remembered, however, that the test gives only a relative
idea of the temperature at which an oil will cloud on long exposure."
Mr. Mory considers the cloud test the most satisfactory for all purposes, being less
subject to conditions.
MAX H. WICKHORST.
(1) For railroad purposes, the clouding test is probably of little service, and probably
what is needed is the flowing test.
(2) I should be in favor of making the cold test on oil as received, without drying or
filtering, as the oil is used in the condition in which it is received.
(3) I do not feel that I can express an opinion with much positiveness, but it would
seem desirable to stir the oil- until solid, so as to avoid any serious segregation.
(4) It would be most convenient to warm up simply by exposing the bottle wrapped
with asbestos or other material to ordinary room temperature; in case of such oils as
tallow, a higher temperature must be used. Whether exposure to room temperature
gives the best or most accurate results, I am not prepared to say.
W. D. RICHARDSON.
(1) (a, b) Both tests are of value depending on the use of the oil.
(2) Dry but not filter sample for cloud test. Dry over flame to 110° C. till free from
moisture.
(3) (a) Yes. (6) Must stand from one-half hour to one hour in freezing mixture to
be thoroughly chilled.
4. (a) Yes. Oils flowing below room temperature, (b) Oils flowing above room
temperature must be warmed up slowly in an air bath with continuous stirring.
J. p. MILLWOOD.
(1) A flowing test only is necessary, as a clouding test will reveal nothing in regard
to its value as a lubricant.
(2) The oil should be examined undried and filtered, as this is the condition in which
it will be used.
(3) It should be left at rest until solid, as the frequency and rapidity of stirring is
liable to introduce a personal equation, and, furthermore, in actual use it is at rest in an
oil cup or oiling can. In order to get concordant results, standing a definite time is
necessary. We find one hour sufficient.
(4) For oils normally liquid, room temperature is sufficient. A bath of warm waller
is necessary for fats and greases.
J. E. WEBER.
(1) Method: Use "flowing test."
(2) Preparation of oil for analysis: Heat the oil to 150° C, but filter only when the
oil is not clear.
31104— No. 105—07—3
34
(3) Method of cooling: Stir the oil until almost solid, then start the test right away.
(4^ Method of melting: Let the oil warm up by placing the bottle in a holder of mag-
nesia asbestos pipe covering. It is very convenient.
ROBEET JOB.
(1) With the ordinary lubricating oil we have never found necessity for the clouding
test, since the result of the cold test gives the desired information as to properties in
service.
(2) If oil contains moistm'e or dirt it is desirable to filter.
(3) As stated above, we find that it is unnecessary to stir until solid; also that the
time of standing after once becoming solid is immaterial.
(4) As to melting, we think it preferable to warm up at room temperature.
E. D. OILAE.
It is apparent that drying a sample before testing does affect the results. The drier
the sample the lower the '"cloud test," and the writer disagrees with Mr. Manns on
this point. The sample should be heated for a few minutes, to a point where it is known
that all crystallized glycerids of the fat, visible or invisible, are melted, somewhat as is
requhed by the New York Produce Exchange method on winter cottonseed oils, but
not to 150° C.
The drying of the sample would create false conditions — that is, if a commercial sam-
ple is tested and clouds at a higher temperature than the guaranty, it matters not to
the consumer whether the sample is wet or dry; the clouding at the undue temperature
remains the same, whereas if the sample were dried, its cloud test might be strictly up
to gi'ade; however, the consumer is obliged to use the oil which does actually cloud at
this temperatm-e for which the chemist's report would be •• O. K." if tested on the dried
sample.
The wi'iter would recommend a flowing or " " cold test "" for lubricating oils and a • ' cloud
test" for edible oils.
Samples should be neither dried nor filtered before testing, for reasons given above.
The writer has not experimented on the time that the sample should remain in the
freezing mixtm'e. but it seems that an hour would be sufficient, especially if a uniform
test is to be adopted.
The melting should be effected in an insulated receiver, such as is described above
for melting at low temperature, or heated in a water bath for higher tests.
It seems probable that a concise "'cloud test" may be devised for winter cottonseed
oil, whereby a given bulk of oil at a given temperatm'e will be placed in an air bath of
given dimensions and of a given temperature, which would be equivalent to the five-
hour test now required by the New York Produce Exchange, and if specifications were
adhered to very closely the time would not enter into the problem.
It is quite apparent that conditions control the results on any of the above tests, and
specifications are necessary as to temperatm'e of cooling baths, volumes, etc.
{R. D. OiLAE made a more complete study of the question, and the following table
gives the results of his work with the cloud test]:
35
special study of the cloud test (Oilar).
Stirring as the oil is cooled.
Not stirring as the oil
is cooled.
Sample.
Tempera-
ture of
clouding.
Tempera-
ture of
bath.
Remarks.
Tempera- Tempera-
ture of ture of
clouding. bath.
°C.
( 33.4
30.6
29.5
29.8-30.2
/ 63.5
i 60.4
/ 78
\ 76.5
°C.
[■Reheated to 150° C
^C. \ ^C.
1
17.6
17.6
14-17. 6
57-59
51-55
75.2
64—68
1 36-37^1 „_„32
3
4
JLower reading due to colder bath
\ 33-34
64-66. 2
59-64
77-78. 8
/
[The figures show that a higher cloud test is obtained by allowing oil to remain
quiet during the test. The following table gives his results on the flowing test]:
Special study of the flowing test (Oilar).
1
stirring as the oil warmed. (Millwood.)
Not stirring as the oil
warmed.
Sample, i
i Flowing
1 test.
1
Tempera-
ture of
bath.
Remarks.
Flowing
test.
Tempera-
ture of
bath.
1 [ 43.5-45
°C.
Room tem-
perature,
do
48.2
61 5
"C.
50
3 40-42
do
(i9 fi
4 82.5-83.5
do....
85.2 100.4
[This study shows that results are much higher when the oil is not stirred as it melts,
and all the data indicate that the methods used must be followed very closely as to
details in order to obtain any satisfactory results.]
A. H. SCHMIDT.
In the Millwood test I found the lard oil to flow at -f-5.5° 0., whereas when placed
in a test tube, immersed in a freezing mixture, and constantly stirred with the ther-
mometer, it still flowed at —3.0° C. The same general results were obtained with the
other oils.
I also believe that the Millwood test is not the correct way of testing a lubricating
oil, as that is really the point at which a frozen or rather solidified oil becomes semi-
fluid, whereas the cold test for a lubricating oil should be to determine at what tem-
perature the oil ceases to fulfil its functioji as a lubricator by congeal'ng. and I found
that to be much lower in most cases than is indicated by the Millwood test.
A. LOWENSTEIN.
(1) It is our opinion that a flowing test should be used for lubricating purposes,
and a cloud test for edible oils.
(2) If the oil when melted is perfectly clear, I do not think that it needs to be
further dried. If when melted it is turbid or contains suspended matter, it should
be filtered through several thicknesses of filter paper, and if still turbid should be
dried by heating to 105° C. until all moisture is removed.
(3) (a) If stirred at all, it should be stirred until it is solid, and the temperature
of the bath should be so low and (6) it should be left "'n the bath so long that there will
be absolutely no question that the oil is solid. I think that the temperature of the
36
bath is here the most important cousideration. and upon it the length of time which
the oil remains in it depends. With the bath at 15° F. below zero one hotir will
suffice, but for higher temperattire a longer time is required, especially for oils of such
low cold test as winter-pressed neat's-foot oil.
In our regular method we do not stir the oil at all, but place it direct in a •"freezer"
at a temperature of —5° F. and allow it to remain overnight. If you Avill examine the
results on tallow oil obtained by Millwood's method and the one employed in this
laboratoiy. you will note a difference of 10°, and the way we accotmt for this is that
in stirring or agitating while chilling, a large amount of air is worked into the oil, and
you get very mtich the same effect as in the agitation of soap or lard or cream. In the
case of lard, for example, one which has a titer of 37.5° C. and which has been agitated
thorotighly will hold up in a warm room better than a lard having a titer of 38.5° which
has been chilled without agitation. In oils containing a smaller percentage of stearin
this difference is not so noticeable, and in the case of neat's-foot oil and lard oil the
results obtained by the two methods check quite well. I do not think that oils of
cold tests over 40° F. need to be stirred while chilling, as there is ample time to render
the mixture homogeneous by stirring when making the flowing test.
(4) (a) I do not think the oU should be allowed to warm up at room temperature.
(b) A bath would be satisfactory, but would have to be different for each kind of oil.
A bath a few degrees lower than the cold test of the oil would, I think, be satisfactory.
In the case of oils of verj^ low cold test, these should be run Yery quickly after remov-
ing from the freezing mixttu'e. othei-wise the oil will melt next to the walls of the
bottle and make it difficult to get the proper flowing test.
If the cloud test is used, the bottle should have an air jacket and not be placed
directly in the bath. It is quite difficult in some cases to prevent the oil from freezing
to the sides and bottom of the vessel.
COMMEXTS OF THE ReFEREE.
The results obtained as shown in the table are certainly very divergent, considering
the details given in the instructions, and show that, as the methods have been
described, some essential directions are lacking or the personal equation is too lai^e
to make them of any value, when two men. following exactly the same method, can
differ 31° F. in determiniag the point at which an oil flows from one end of a bottle to
the other. That the personal equation does enter largely is shown by one analyst
obtaining the lowest restdts in three out of the fotir samples. The same tendency
may be noted in the highest results.
The following ctirious point is also brought out: The temperature at which the
cloud forms in cooling an oil down in most of the oils is several degrees below that
at which the oil flows when being melted after cooling. That is, under the methods
employed the oil is liqtiid and clear below the temperature at which it flows on
being melted. For example, lard oil clouds at 30.2° F., congeals at 19.4° F.. and on
remelting flows at 42° F.
It is, however, true with nearly aU substances that the cri,*stallizing point is below
the melting point, but with the grease oil this is not the case, the ciystallizing point
being above the flowing point.
Titne is perhaps the most important factor. As has been shown in the determina-
tion of melting points of fats, it is necessary to allow them to stand a number of hours
after they have been melted and recooled before the correct melting point is obtained,
and the same fact is doubtless true in this determination, and also in the melting point
it is necessarj^ to allow the temperatiu'e to rise verj^ slowly in order to obtain an>n;hing
like the true melting point, and if this is true in melting points of pure compounds it
certainly must be true in such a determination as this.
It seems to the referee that before the flowing test can be of any value some method
of warming up the frozen fat must be employed, such as suggested by Lowenstein,
37
who takes the sample from the freezer and places it in a bath at about 40° for fifteen
minutes, so as to allow it to soften, then stirs until it flows. This procedure, of course,
could not be employed with oils with a flowing test below 40° F. Some arrangement
of this kind, varied in temperature somewhat for the different oils, seems necessary
from the results obtained this year. This might be accomplished as follows:
After the oil has been frozen (and the temperature of the freezing bath should be well
below that of the melting point of the oil so that it is frozen hard) the sample is placed
in a water bath 5° to 10° below the flowing test and allowed to warm up to that tem-
perature. Then finish its determination as now directed.
As to the clouding test, it seems to be shown that there is no relation between the
flowing point of an oil and the temperature at which an oil clouds, and as the former is
the question in lubricating oils, it would seem that the cloud test is hardly applicable.
The consensus of opinion of the collaborators seems to be that this test is of value
only on salad oils. The results obtained with the cloud test on the various oils were
more uniform, as given in the table, showing that the method used gives comparable
results, although the maximum and minimum show a wide variation, due to the freez-
ing of the oil on the sides of the bottle. This might be obviated, as suggested by
Mr. Lowenstein, by surrounding the bottle with an air jacket, so that the walls of the
bottle do not come in direct contact with the cold bath.
Summary.
The work on fats and oils this year, is a continuation of that begun last year on the
cold test and is of a preliminary nature. Considerable cooperative work has been
done, fourteen men sending in reports on the samples. The results, while not agreeing
in a satisfactory way, have brought out a number of important points, and will enable
the referee to continue the work with a clearer idea of what is necessary. The results
showed that there were two entirely different tests known as the "cold test," and
names have been given to these which will in the future distinguish them from each
other.
It was clearly shown that the cloud test and the flowing test are dependent on
different factors and can not be used for the same purpose, and that the flowing test is
the one which is of value in lubricating oils, while the cloud test is more applicable to
salad oils.
The results also showed the need of more definite, detailed directions for the flowing
test, indicating that it will be necessary to warm up the fats in a bath in which the
temperature can be controlled. It was also shown that with the cloud test there is
need of a bath to prevent the overcooling of the sides of the tube.
The work of the year has been very valual)le, and it seems probable that in another
year a method can be devised which will give satisfaction.
EEPORT ON DAIEY PEODUOTS.
By Albert E. Leach, Referee.
During the past year a large number of samples of milk have been examined in the
laboratory of food and drug inspection of the Massachusetts State board of health l)y
the provisional method for the detection of added water from a refractometric examina-
tion of the milk serum. The method has been of great service in routine work for
distinguishing between milk to which water has been fraudulently added and milk
that simply is below the legal standard as it comes from the cow.
In view of the fact that according to law $50 is the minimum fine in Massachusetts,
if the milk can be proved in court to be actually watered, while the fine for milk simply
below the standard may be as low as the judge sees fit to impose, the importance of
being able to distinguish between the two classes and to prove the presence of added
38
water is evident. More than 40 court cases have been tried in our State courts during
the past year in which the complaint for added water was based on the pro^-isional
refractonietric method, and in at least 95 per cent of these cases con^'iction was secured.
The accompam-ing chart has been prepared to show graphically the results of the
determination of solids not fat and of the refraction of the serum of 141 samples of milk
all below 12 per cent in total solids. The lower curves show the variation from sample
to sample in the percentage of solids not fat, while the upper cur^-es show the refraction
of the milk serum in the same samples. The samples are arranged in the order of the
solids not fat, the refract ometric reading corresponding to the solids not fat in each
sample being readily apparent as it is placed along the same ordinate in the chart.
Af/L/r FR/i UDUL ENTLY\
lA/JHTERED i
M/LM BEL014/ LEG/IL ST/IND/IRD
^/
39
38
•57
36
35
3^
33
32
/O
9
8
7
6
5
1
Fig. 1.— Comparison of refractometric examination oi milt serum and the determination of solid:
not fat for detection of added wat«r.
The first 40 samples of milk were pronounced fraudulently watered, since all stand
in refraction of serum below 39°. as shown by the upper cwcxe. The rest of these
samples, even though they all were below the legal standard of total solids, can not be
declared fraudulently watered.
If we were to determine the watered samples from the solids not fat alone, assuming
7.3 as the mininnm-i solids not fat for a pure milk, we could only condemn 29 of the
samples as being fraudulently watered, thus allowing at least 11 samples to escape
detection. It is also apparent from a comparison of the two curves that there is no
fixed proportion between the solids not fat and the refraction of the serum, since in
some cases a milk low in one constant is found to be high in the other.
39
EEPORT ON CONDIMENTS OTHEE THAN SPICES.
" By R. E. DooLiTTLE, Associate Referee.
In presenting the report on condiments other than spices there is but little to offer
in the way of original investigations. The referee, finding that no methods for the
examination of this class of products had ever been presented to the association, has
outlined certain methods which in his experience are of value in judging the character
and purity of products coming under this subdivision. These methods are not
restricted to products of this class but are, at least for the most part, methods appli-
cable to other products, especially those closely allied to condiments other than spices.
In fact, this subdivision overlaps some of the other subjects for which methods
have already been adopted provisionally. In this classification, therefore, only such
products have been included as are not generally included under any of the other
subdivisions.
The referee recommends that these methods be adopted provisionally but that in
the printing of the same they be given by reference wherever such methods appear
under other subjects.
Methods for the Analysts of Condiments other than Spices.
(1) general discussion.
Under this head are included those mixed and prepared products used for the pur-
pose of seasoning or giving relish to food. The various table sauces, catchups, horse-
radish, chutneys, and soys may be enumerated as examples of the principal products
designated by this term. These preparations often vary widely in composition, as
a various number of substances are used in their manufacture according to recipes that
have been found to produce a characteristic flavor or pungency. In the examination
of these products it may be assumed that no substance should be present that has no
value as a food or flavoring material. Inert substances, known commercially as fillers,
should be regarded as adulterants. Preservatives and coloring matters are the other
adulterants most commonly present.
(2) MICROSCOPIC EXAMINATION.
Submit all samples containing solid substances to a microscopic examination for
identification of these ingredients.
(3) PREPARATION OP SAMPLE.
Insure a uniform mixture by thorough stirring or shaking. Products containing
solid substances should be reduced to a pulp by passing through a food chopper or
grinding in a mortar.
(4) DETERMINATION OF SOLIDS.
Weigh 5 grams of the substance into a large flat-bottomed platinum dish, evaporate
on steam bath to apparent dryness and then in oven at 100° C. to constant weight.
(5) DETERMINATION OF ASH.
Thoroughly char the residue from determination of total solids at as low a heat ao
possible, extract with hot water, filter, and wash. Return the filter paper holding the
insoluble matter to the dish and thoroughly ignite; add the soluble portion, evaporate
to dryness, and ignite carefully at low redness. Cool in desiccator and weigh.
40
(61 DETERMIXATIOX OF CHXOEIX.
Dissolve the ash in hot water, cool, and dilute to 250 cc. Titrate 50 cc of solution
with tenth-normal silver nitrate, using potassium chromate as indicator. «
(7) DETEEMIXATIOX OF TOTAL ACIDITY.
Dilute 5 grams of the material to 500 cc, thoroughly mix and titrate 100 cc of the
diluted solution with tenth-normal soditim hydroxid. using phenolphthalein as
indicator.
(8) DETERMIXATIOX OF VOLATILE ACIDS.
To 10 grams of the material add 50 cc of water and distil in a current of steam until
250 cc pass over. Titrate the distillate with tenth-normal sodium hydroxid. using
phenolphthalein as indicator.
U DETERMIXATIOX OF XITROGEX.
Determine nitrogen on 5-gram sample by Kjeldahl metliod.&
(10) POLARIZATION.
Dilute 26.048 grams of the sample to about 150 cc in a 200 cc sugar flast. add an
excess of lead subacetate, make to mark, thoroughly mix, filter, and polarize in a
200 mm tube, observing the temperature at which the reading is made. Invert 50
cc of the filtrate in a 100 cc sugar flask, using 5 cc of hydrochloric acid and heating
to 68 C. in 15 minutes, nearly neutralize with dilute sodium hydroxid. cool, make
to mark, and polarize in 200 mm tube. Correct to normal weight by multipl^'ing
the direct reading by 2 and the invert reading by 4. Both readings should be made
at or near 20° C.
Determination of cane sugar.
(O) FEOM POLABIZATIOSr.
From the direct and iuA^ert readings calculate the cane sugar by the modified
Clei^et ' s formula ,
g_100 (g— &)
142.66— I
(6) Feom Inceease rtf Reductsg Sugars.
When only small percentages of cane sugar are present it may be best determined
from the increase in reducing sugars after inversion. The increase in per cent multi-
plied by 0.95 equals percentage of cane sugar.
(11) DETERMIXATIOX OP EEDUCESTG SUGAKS BEFORE IXVERSIOX.
Transfer 10 cc of solution prepared for direct polarization to a 250 cc flask; add
sufficient sodium sulphate to precipitate excess of lead, make to mark, mix, and filter
through a dry filter. Determine reducing sugars by official methods for determina-
tion 01 invert sugar, c
(12) DETERMIXATIOX OF REDL^CDsG SUGARS AFTER IXVERSIOX.
Transfer 5 cc of invert solution prepared for polarization to a 250 cc flask, make to
mark, and determine reducing sugars by official method for determination of invert
sugar, c
« Sutton's Volumetric Analysis, 9th edition, page 42.
&U. S. Dept. of Agr.. Bureau of Chemistry. Bui. 46, p. 14.
cBAd., p. 33.
41
(13) DETERMINATION OF HEAVY METALS.
Proceed as directed under ' ' VII. Canned vegetables," Bulletin 65, page 53, 10 (b).
(14) DETECTION OF COLORING MATTER.
Proceed as divected under " XVII. Coloring matter," Bulletin 65, page 111.
(15) DETECTION OF PRESERVATIVES.
Proceed as directed under " Food Preservatives," Circular 28.
REPOKT ON TEA AND OOrPEE.
By C. T). Howard, Associate Referee.
Introduction.
The time available for this study has been unexpectedly limited this year, no
cooperative work having been attempted, and as far as the improvement of methods
is concerned the results attained have been largely negative. The work has been
confined principally to a study of methods for the estimation of caffetannic acid.
Of late a number of brands of coffee have made their appearance concerning which
the claim is made that the "tannin" has been removed. The following analyses,
made at the New Hampshire laboratory of hygiene, show the relative composition
of three of these so-called tanninless coffees, also of ordinary coffee and of coffee
chaff:
Comparison of analyses of tanninless coffees, coffee, and coffee chaff.
Sample.
Water.
Ash.
Fat.
Fiber.
Caffeine.
Caffetan-
nic acid
(Krug's
method).
No. 1]
No. 2>Tanninless coffees . . .
Per cent.
[ 2.70
\ 2.70
[ 2.26
3.13
2.60
Per cent.
4.10
4.05
3.61
4.13
5.65
Per cent.
13.18
14,12
12.55
14.10
9.30
Per cent.
18.46
15.70
22.70
15 50
26.50
Per cent.
1.17
L33
.87
L29
.40
Per cent.
10.76
11.04
No. 3j
Java and Mocha
7.61
11.17
Coffee chafif
5.98
In two of these cases the manufacturers base their claim for the ' ' complete removal
of tannic acid " upon the elimination, during the process of grinding, of the chaff or
skin occurring on the surface and, to some extent, enmeshed in the grooye of the coffee
bean. The fallacy of such a claim, however, is evident from the composition of tha
chaff. Sample No. 3 represents a coffee of a different type, the claim being that by
certain extractive processes "all of the tannin and caffeine" have been removed.
As there are now several processes of tliis nature and as the popular demand for
"hygienic" foods is likely to add to the number of brands of such products now upon
the market, the desirability of accurate and fairly rapid methods for the estimation
of caffetannic acid and caffeine is obvious.
Caffetannic Acid.
CONSTITUTION.
The exact constitution of caffetannic acid still seems to be a matter of some doubt.
Allen (Commercial Organic Analysis) gives the empirical formula as Cj4 Hjg O7; in
Smith's translation of Pichter's Organic Chemistry it is represented as C30 Hjg Oig,
while Krug bases a factor on the formula advanced by Hlasiwetz,^ viz, C^g Hjg Og.
« Ann. Chem. Pharm., 1867, 142 : 230.
42
The later investigations of Cazeneiive and Haddon.a represent this body as being a
yCH=CH. COOH
diglucoside ofdioxycinammic acid. & having the composition Cgllg— 0. C6Hii05
They find that the acid gives a crystalline osazone of yellow needles melting at 180°,
very slightly soluble in alcohol, of the following composition:
yCH=CH. COOH
CeH3-0. CeHA(X-XH. CeHs)^
\0 . C,H,03( X— XH. CeH^).
^^^lile it would seem that the Cazeneuve and Haddon formula — indicating a di-
rather than a mono-glucoside — is probably correct, yet so far as can be learned it has
not as yet been verified by other investigators: unforttmately no time was available
for any study in this connection.
The form of combination in which caffetannic acid exists natiually has not as yet
been definitely established. Bell (Chemistry of Foods) designates it as a "'chloroge-
nate (caffetannate) of potash and caffeine." This view is probably based upon the
work of Payen c published in 1849. Allen (Commercial Organic Analysis) states that
caffetannic acid occurs ''probably as a calcium of magnesium salt." That the latter
form of combination, or one similar, is the true one would seem to be indicated in
some degree by the comparatively recent investigations of Trillich and Gockel d in
connection with a proposed method for the estimation of this body based upon extrac-
tion with ether and benzol in the presence of an acid.
As prepared by precipitating an alcoholic infusion of coffee with lead acetate, fil-
tering, and decomposing the lead precipitate with hydrogen sulphid, caffetannic acid
forms a brownish sirup-like mass, having a slightly acid and astringent taste. The
lead salt on drying becomes hard and brittle and is apparently somewhat hygroscopic,
BEHAVIOR TOWARD REAGEXTS.
On boiling caffetannic acid with dilute acids, glucose and caffeic acid (di-oxyci-
nammic acid) are formed. Caffeic acid also results on heating the solution with
caustic alkali. Fusion with caustic potash yields pyrocatechuic and acetic acids.
Caffetannic acid precipitates solution of gelatin but slightly, and this only in the pres-
ence of a large excess of the reagent, thus affording a marked contrast with the behavior
of tea tannin.
Iron alum gives a gi-eenish black precipitate or color: bromin and iodin gii'e light,
finely flocculent precipitates of a yellowish color. Copper acetate affords a greenish
precipitate, filtercdwith difficulty; ammoniacal copper-sulphate solution a bulky pre-
cipitate, the greater part of which is easily soluble in a slight excess of ammonia.
Lead acetate jdelds a voluminous precipitate, calcium and barium hydroxids
slighter ones; the latter tend to become slimy on the filter and wash with difficulty.
An ammoniacal solution of potassium ferricyanid fails to produce the characteristic
blood-red color afforded by dilute solutions of the ordinary tannins, the tint produced
being dull brownish red. The volumetric procedure of Fletcher and Allen employed
for estimating the tannin of tea by precipitation with standard lead acetate solution
is therefore not applicable to coffee.
Unlike that of tea. the tannin of coft'ee gives no precipitate whatever with neutral or
ammoniacal zinc solutions. Certain alkaloidal salts (quinin and cinchonin especially)
a Comptes Rendus, 1897. 1^-^:1458.
& See also the- work of Kunz-Krause, Ber. d. chem, Ges., 1897, 30: 1617.
cAnn. chim. Phys. (3t 16: 108.
dZts. Nahi', Genussm., 1898, p. 101.
43
produce a somewhat erratic precipitation of caffetannic acid, but although some exper-
iments were made using various indicators, this fact does not seem to be applicable
as a basis for a quantitative method.
SOLUBILITY OF LEAD CAFFETANN ATE.
In view of the fact that four of the five methods described below for the estimation
of caffetannic acid involve precipitation as the lead salt, the following facts were
determined relative to the solubility of this compound:
An infusion of roasted coffee, prepared by boiling with distilled water, exhibits a
slightly acid reaction. If to such a hot infusion a perfectly neutral solution of lead
acetate be added, the acidity of the filtrate from the resulting lead caffetannate will
be found to be appreciably increased. Because lead caffetannate is slightly soluble
in hot dilute acids it is therefore evident that precipitation by neutral lead acetate
can never be quite complete, although if proper precaution be taken it is probable
that the loss from this cause is inappreciable. Hot, moderately strong acid dissolves
lead caffetannate completely. This salt is also very appreciably soluble in a large
excess of neutral lead acetate. If basic lead acetate (avoiding excess) be employed
as the precipitant, the resulting filtrate will be found to give no reaction for tannins
with ferric acetate.
Lead caffetannate is slightly soluble in boiling distilled water, and in boiling 10 per
cent alcohol. The perfectly clear filtrate obtained from boiling with 20 per cent
alcohol gives no reaction for tannins with ferric acetate. With the use of stronger
alcohol, however, a clear filtrate is more readily obtainable.
METHODS OF ESTIMATION.
Various methods for the quantitative determination of caffetannic acid have been
proposed, no two of which give the same results, and all of which are in some degree
fallacious. The chief difficulty has been in the fact that the true nature of this body
has not until recently been understood, and that even now its exact constitution is
not proven beyond question. In the past at least one method has been applied to
some extent, based upon the assumption that tea tannin and coffee tannin gave simi-
lar reactions.
BelVs, or Payen's Method.
Five grams of coffee are exhausted with alcohol. The alcohol is evaporated and to
the aqueous residue subacetate of lead is added. The precipitate is thrown on a filter,
washed, decomposed with sulphuretted hydrogen. The filtrate from the plumbic
sulphid is evaporated to dryness, when the caffetannic acid is obtained as a yellowish
brittle mass.
By this method Trillich and Gockel found 5.32 and 3 per cent, respectively, of caf-
fetannic acid in raw and unroasted coffee. Two s.erious defects in the above method,
pointed out by Trillich and Gockel, are that every evaporation of the caffetannic acid
extract results in a considerable proportion of the latter acid being rendered insoluble,
and also that it is impossible by the ordinary procedure to completely decompose the
lead precipitate with sulphuretted hydrogen on account of the tendency to occlusion
of masses of this precipitate by the lead sulphid.
Proctor's Modification of Low enthaV s Method.
This method, official for estimating the tannin of tea, has also been applied to some
extent in the case of coffee, more especially for the sake of obtaining comparative
values on different brands. That the results obtained possess only relative values
is evident from the behavior of coffee extract with gelatin, and also because of the fact
that the ordinary tannic acid values for the permanganate solution have no application
in this case.
44
Through the courtesy of a Boston coffee house data obtained in 1906 by the analyses
of t\i^enty samples of roasted coffee were made available. The aA'erage percentage of
tannin, estimated by the Lowenthal process, was found to be 4.63.
Krug's Method.
Ti'eat 2 grams of the coffee with 10 cc of water and digest for 36 hours; add 25 cc of
90 per cent alcohol and digest 24 hours more; filter, and wash with 90 per cent alcohol.
The filtrate contains tannin, caffein, color, and fat. Heat the filtrate to the boiling
point and add a saturated solution of lead acetate. If this is carefully done, a caffe-
tannate of lead will be precipitated containing 49 per cent of lead. As soon as the
precipitate has become flocculent, collect on a tared filter, wash with 90 per cent
alcohol until free from lead, wash with ether, dry, and weigh. According to Krug
the precipitate has the composition Pbg (C15H15 03)2. Weight of precipitate multi-
plied by 0.51597 gives the weight of caffetannic acid.
By this process Krug found 10.88 per cent of caffetannic acid in a sample of unroasted
Java. Trillich and Gockel, working on an unroasted New Granada coffee, obtained
the figures 11.12 and 11.50, while a roasted coffee of this variety gave 10.68 and 11.32
per cent. Results obtained by the writer using this method on roasted coffees varied
from 10.65 to 11.17 per cent, duplicate determinations on a sample of Java and Mocha
blend giving 10.65, 10.69, and 11.02 per cent. Some difficulty was experienced in
securing a constant weight for the lead precipitate.
The principal objection to this method is its tediousness, the preliminary digestion
consuming a large amount of time and the washing of the lead precipitate being a
very slow and troublesome process.
Method of Trillich and Gochel.f^
Boil 3 grams of coffee one-half hour with water, filter, and repeat this treatment on
the residue three times. The united filtrates are made up to 1,000 cc. To 400 cc add
1 cc of basic lead acetate solution and allow to stand over night. Filter, wash, decom-
pose the precipitate with sulphuretted hydrogen, filter from lead sulphid, evaporate
to dryness and weigh.
Trillich and Gockel obtained by this method 10.88 and 11.37 per cent on raw and
7.99 and 8.60 per cent on roasted coffee. These authors also attempted to base a method
upon extraction with ether in the presence of hydrochloric acid, but found that boiling
with the latter acid occasioned some destruction of the tannin, while phosphoric acid
did not appear to be capable of liberating caffetannic acid from the bases with which
it is combined.
They conclude, as a result of their investigations, that the structure of caffetannic
acid is not altered by the roasting process; also that all methods based upon precipita-
tion with lead acetate are in some degree fallacious in that such precipitate appears
not to consist wholly of caffetannate.
Hydrolytic method.
Some time was given to a study of a method involving hydrolysis of the glucosid
by boiling with hydrochloric acid and subsequent estimation of the resulting glucose.
Five grams of coffee were boiled one-half hour with 100 cc of water and the extract
decanted through a filter. This treatment was repeated several times and the filtrates
made up to 500 cc. From. 200 to 400 cc, equivalent to 2 to 4 grams of coffee, were pre-
cipitated with lead acetate, the precipitate washed sufficiently to remove the small
proportion of sugars present and then submitted to hydrolysis by heating in a water
bath with hydrochloric acid. The reducing sugars were then estimated by Allihn's
method.
aZts. Nahr. Genussm., 1898, p. 101.
45
In the following table the percentages of reducing sugars, considered as CgHi^Og,
are calculated to caffetannic acid for both the mono- and di-glucosid formulas attrib-
uted to this body.
Estimations of caffetannic acid by the hydrolytic method usimj varying strcnyths of hydro-
chloric acid.
Caffetannic acid. 1
Strength
Time.
Cupric oxid
Reducing
of acid.
in 0.5 gram.
sugars.
For C21.
For Ci5.
Per cent.
Hours.
Per cent.
Per cent.
Per cent.
2h
3
0. 0480
4.02
5.63
7.28
2i
6
.0647
5.36
7.50
9.71
5
3
.0600
4.98
6.97
9.01
5
4
.0625
5.38
7.53
9.74
5
6
.0795
6.52
9.13
11.80
5
8
.0890
7.28
10.19
13.18
n
3
.0840
6.88
9.63
12.45
"'h
4
.0795
6.52
9.13
11.80
il
5
.0770
6.32
8.85
11.44
l\
5
.0737
6.06
8.48
10.97
6
.0790
6.48
9.07
11.73
71
8
.0800
6.56
9.18
11.87
10^
3
.0755
6.20
8.68
11.22
10
6
.0597
4.87
6.82
8.81
It is evident that this glucosid offers considerable resistance to decomposition.
The strength of hydrochloric acid ordinarily employed for hydrolysis is insufficient,
acid of between 5 and 10 per cent affording the maximum yield of sugar. Recognizing
that the Krug method gives results necessarily somewhat too high, the above values
would appear to point to the correctness of the Cazeneuve and Haddon formula,
especially as under the conditions most favorable to complete hydrolysis (not yet
ascertained), the proportion of caffetannic acid, assuming the formula CigHigOg,
would be materially in excess of that actually afforded by the Krug method.
This method is much less tedious than that of Krug, and has the advantage that the
results obtained are directly referable to a definite formula for caffetannic acid. Not
only are there differences of opinion as regards the composition of the lead salt, but it
is certain that the percentage of lead in the latter varies with the conditions under
which the precipitation is effected. On the other hand, the hydrolysis of the caffe-
tannic acid, as carried out by the writer, proceeds somewhat erratically.
Recommendations.
It is recommended:
(1) That the methods here outlined for the estimation of caffetannic acid be sub-
mitted for cooperative study by the members of this association.
(2) That the Gomberg method for the determination of caffein be subjected to trial.
This method can be logically carried out in connection with the caffetannic estimation,
and is essentially as follows: The filtered extract, prepared by boiling the material
witli water, is precipitated by lead acetate and the precipitate filtered off and
thoroughly washed with water. The excess of lead is removed from the filtrate by
sulphuretted hydrogen and the latter expelled by boiling. An excess of a standard
potassium iodid solution of iodin is added, the precipitate of the alkaloid with iodin
filtered out, and the remaining iodin estimated by titration with sodium thiosulphate.
46
EEYIEW or METHODS POE ANALYSIS OF TEA.
By R. E. DooLiTTLE and F. 0. Woodruff.
The authors being called upon during the past year to examine a number of samples
of tea. have had occasion to study some of the many methods given in the literature
for the chemical analysis of this product. As the results of these studies may be of
use to the referee on the subject of tea. the following report is submitted.
The samples used in these investigations were the standards of 1905-6 employed by
the Treasury Department for the examination of imported teas, and were of the follow-
ing varieties:
Varieties of tea examined.
Variety
Labo-
ratory
No.
Oolong
Foochow oolong
North China congou
South China congou
India
Ping Suey
Varietv.
Labo-
ratory
No.'
Country green 7
Pan-fired Japan*. 8
Basket-fired Japan , 9
Japan dust, or farmings | 10
Capers i 11
(1) Determixatiox of Water Extract.
In the study of the methods for determination of water extract the method of Krauch.
reported by the referee on tea and coffee at the twenty-first annual meeting of the
association, was first considered.
Method of Krauch o. — Treat 20 grains of the tea with 400 cc of water and heat on a
boiling water bath for six hours. Filter through a tared filter, wash with water until
the filtrate measures 1.000 cc dry. and weigh the residue. The water-soluble sub-
stance is determined by the difference.
The method as given is rather A'ague in some respects, and after some preliminary
experiments the samples were run as follows :
Twenty grams of the tea were weighed into a liter flask. 400 cc of hot water added,
the flask heated on steam bath for six hours, the A"ulume of water being kept at 400 cc by
addition of hot water from time to time. The contents of the flask were then filtered
through a tared filter and washed with hot water until filtrate measured 1,000 cc. The
residue was dried by transferring paper and contents-to tared aluminium dish and heat-
ing to constant weight in steam oven.
The results obtaincvd by this method were as follows:
Determination £>f water extract by th£ Krauch method, moaified.
Laboratory No. ■
Per cent extracted
in duplicate.
i Laboratory No.
i
Per cent extracted
in duplicate.
1 1 32. 26 40. .56
_
.30. 44
31. 81
32.18
29.93
31.44
^39.42 ,
30.27
34.76
34.16
35.14
1
2 ' 30.55 33.26
8
3. . . .31.60 .3.5.87
9
4
5
6
27. 72 25. 44
34. 12 35. 13
32. 67 33. 32
10
11
Triplicate. 39. 87 per cent.
Several difficulties were encountered in the manipulation of the method, which was
found to be long and tedious, requiring in some cases, with green teas, two days to secure
the 1.000 cc of filtrate. The large amount of sample used made it almost impossible
a L'. S. Dept. Agr., Bureau of Chemistry. Bui. 90. p.
3d edition, p. 1057.
Konig, Zts. Xahi'. Genussm.
47
to dry the residue to constant weight, and the size of the filter paper required undoubt-
edly gave a large error from change in weight. It was not possible to completely
extract so great a quantity of tea with the volume of water used. Duplicate determi-
nations did not check sufficiently close.
As it was found that the principal difficulties in the manipulation of the Krauch
method arose from the large quantity of sample used for the determination, it was
decided to try smaller quantities. From our own experiments and a review of the
methods given in the literature 2 grams was decided upon as a sufficient quantity for
extraction.
To determine the length of time necessary to secure complete extraction of the 2
grams of tea, sample No. 1 was extracted by boiling with 200 cc of water for various
lengths of time, washing the residues until the total filtrate measured 500 cc. Results
obtained were as follows:
Results obtained by varying the time of extraction.
Per
Per
Per
Time.
cent ex-
Time.
cent ex-
Time.
cent ex-
tracted.
tracted.
tracted.
Minutes.
Minutes.
Minutes.
5
42.47
25
44.53
45
44.78
10
42.51
30
43.82
50
44.54
15
43.32
35
43.91
55
a 45. 62
20
44.21
40
44.06
60
6 45. 67
a Duplicate, 45.87 per cent.
b Duplicate, 45.07; triplicate, 46.50 per cent.
To determine the volume of water required for extraction, 2-gram samples of No. 1
were boiled for 30 minutes with varying amounts of water, all being filtered and washed
with hot water until filtrate measured 500 cc. Results were as follows:
Results obtained by varying the amounts of uater used.
Water
used.
Per
cent ex-
tracted.
Water
used.
Per
cent ex-
tracted.
cc.
100
125
150
175.
44.47
43.82
43.81
44.54
cc.
200
250
300
44.10
44.50
44.10
The matter of weighing the residue was considered, whether or not it is advisable to
weigh separately or with the filter paper. Duplicate samples were run, weighing the
residues in the first instance with filter by means of a weighing bottle, and in the second
by removing the residue to a tared dish and weighing separately. Results were as
follows:
48
Tuo methods of ueiyhing residue compared.
I Amount extracted. \
-
Amount extracted.
\ Residue
Labo- ' and paper
ratory ^ ^vith ad-
No. 1 hering par-
! tides sepa-
rately de-
termined.
Residue
and paper
weighed
together
in weigh-
ing bottle.
Labo-
ratory
No.
Residue i -r^^vIiip
I Per cent.
1 i 0 41.65
2 42.01
3 38. 71
4 46. 35
5 44.62
6 36. 97
Per cent, i
6 44.16 1
43.05
36. 71
38.62
44.92
7
8
9
10
Per cent.
C39.87
38.12
41.24
Per cent.
d 40. 16
4L13
41.27
11
43.67
a Duplicate, 41.71; triplicate. 41.51.
b Duplicate, 43.82.
c Duplicate, 41.47: triplicate, 39.42.
d Duplicate, 41.25.
As a result of these studies the following method is suggested for the determination of
extract in tea:
To 2 gi-ams of the tea in a 500 cc Erlenmeyer flasli add 200 cc of hot water and boil
with low flame for one hour. The flask should be closed with a rubber stopper through
which passes a glass tube 18 inches long for condenser. The loss from evaporation
should also be replaced from time to time by addition of hot water. Filter through a
tared filter and wash the residue until the filtrate measiu'es 500 cc. stirring the contents
of the filter thi'oughout the process to facilitate the filtering. Reserve filtrate and
washings for determination of tannin and theine. Dry the filter paper and residue in
its funnel in the steam oven until the excess of water is removed, transfer paper and
contents to tared weighing bottle and dry to constant weight at 100° C.
i2i Determixatiox of Taxxix.
Tannin was determined on the eleven samples by the Proctor modification of Lowen-
thal's method a with the following results:
Tannin determinations i Proctor-Lou enthal methods
Labo-
ratory
No.
Per cent of tannin
found in duplicate.
If^c.^ Per cent of tannin
Xo ^°^^ ^° duplicate.
1
2
3
4
5
6
8.42
9.62
6.61
7.56
8.76
7.39
7.89
9.45
6.36
7
8
9
10
! 11
i
8.42
6.69
7.04
6.36
9.27
5.33
7.21
6.87
8.58
9.27
5.33
To avoid the necessity of making a separate infusion for the determination of tannin
the experiment was made of using an aliquot portion of the filtrate from the determina-
tion of water extract. The determination was made as follows:
The filtrate from the determination of water extract was made to 500 cc. To 25 cc of
this volume 25 cc of indigo carmine solution and about 750 cc of water were added and
the whole titrated with standard permanganate solution. Two hundred and fifty cc of
the solution was next mixed with 50 cc of gelatin solution. 100 cc of salt solution, and 10
grams of kaolin added and the whole shaken for five minutes. This was decanted
through a filter. To 40 cc of the filtrate 25 cc of the indigo carmine solution and 750 cc
of water were added and titrated as before.
aU. S. Dept. Agr.. Bureau of Chemistry Bill. 90. p. 39.
49
The following results were obtained, which for the purpose of comparison are here ,
tabulated with the results obtained by the Proctor modification of Lowenthal' s method :
Comparison of tannin results by the Proctor modification of the Lowenthal method and
the authof s modification.
Per cent of tannin.
Labo-
Per cent of tannin.
Labo-
ratory
Experi-
Lowen-
ratory
Experi-
Lowen-
No.
mental
thal
No.
mental
thal
method.
method.
method.
method.
1
8.76
8.42
7
6.52
8.42
3
6.76
6.61
9
9.00
7.04
4
7.44
7.56
10
6.92
6.36
5
9.18
8.76
11
9.52
9.27
6
7.76
7.39
To further check the methods, duplicate determinations were made on two samples.
Results obtained were as follows :
(a) First determination, 8.07 per cent; duplicate, 8.42 per cent. (6) First determi-
nation, 8.76 per cent; duplicate, 8.93 per cent. Corresponding to a difference in
titration of 0.2 cc and 0.1 cc, respectively.
The results being as satisfactory when the filtrate from the determination of water
extract was used, which is a saving of much time, the authors would suggest the follow-
ing modification in manipulation of the Lowenthal method : «
Cool and make to 500 cc the filtrate from the determination of water extract. To 25
cc add 25 cc indigo carmine solution and about 750 cc water, and titrate with standard
permanganate solution, beginning with 1 drop at a time for 20 drops, then 1 cc at a
time till color is a light green, then very carefully until a faint pink appears around edge
of fluid. Number of cubic centimeters used = a.
Mix 250 cc clear filtrate with 50 cc gelatin solution, 100 cc salt solution, cork, add 10
grams kaolin, and shake five minutes; decant, then filter. To 40 cc of the filtrate add
25 cc indigo solution, 750 cc water, and titrate as above. Number of cubic centimeters
of permanganate used= h. a — 6 = c =^ permanganate (KMn04) required to oxidize
the tannin; 0.04157 gram of tannin = 0.063 gram of oxalic acid.
(3) Determination of Theine.
Theine was determined on one sample of tea by the following methods with the fol-
lowing results:
Determinations of theine by different methods.
•
Method.
Per cent of theine
in duplicate.
Dvorkowitsch a
2.56
2.38
1.51
2.41
Stahlschmidt b
2.28
Commaille c
a Bureau of Chemistry Bulletin 90, p. 39; also Konig, Zts. Nahr. Genussm., 4th ed., p. 1010.
b Leach, Food Inspection Analysis, p. 281.
c Blyth, Foods: Their Composition and Analysis, 5th ed., p. 334.
The experiment was then made of determining the theine by the Dvorkowitsch
method on an aliquot portion of filtrate from water extract determination. A deter-
mination on the same sample (No. 4) as was used for the above determinations gave
2.62 per cent of theine. The samples run by this modification gave the following
results:
For reagents and standard solutions see Bureau of Chemistry Bulletin 90,
31104— No, 105—07 4
). 39,
50
Determinatio-ns of theine by a modification of tit e DvorJ:ou:itsch method.
Labor-
Melting
Labor-
Melting
atorr
Theine.
point 01
atorv
Theme.
point of
Xo.
crystals.
Xo.
crystals.
Per cent.
oC.
s
Per cent.
or.
1
2. 874
226 1
7 i
2.070
232
••)
2.3%
229
8 1
1.994
231
3
2.5^
234
9
2.032
228
4
2.620
231
10 i
l.So6
233
.-.
3-270
22S
11
1.S9S
231
n
1.716
The authors Trould suggest the folloTring modification of the Dvorkowitsch method
for determination of theine:
Extract 225 cc of the filtrate obtained from determination of water extract made
to 500 cc in a separators* funnel with petrolic ether to remove fat. To the fat-free
portion add 50 cc of a 4 per cent barium hydrate solution., shake well, and filter. To
the filtrate add 50 cc of a 20 per cent sodium chlorid solution and extract four times
with 75 cc portions of chloroform, distil off most of the chloroform from, the four
combined portions, and evaporate the remainder in a weighed platinum dish, dr}' at
temperature of steam oven to constant weight.
(4) Deteemixatiox of Moistuee.
The method usually given for the determination of moisture in tea is to dry 2 to 5
grams of the sample in a flat-bottomed dish at lOO'^ C . Doolittle and Ogden determined
the loss of weight in a number of samples of tea by drjdng in an oven at temperature
of boiling water and by dr^-ing in a current of hydrogen a at temperature of boiling
water. The following results were obtained:
Comparisons of different methods of drying.
Labor-
atorr
Xo.'
Varietv of tea.
Loss in
oven at
temx)era-
ture of
boiling
water.
Loss in
air oven
at 110°.
Loss in
cnrrent
of hydro-
gen.
r. S- standards 1205-6.
Pormosii
Foooho^ oolong
Xorth China congou. . .
South China congou
Ladia
Ping Suey
Country green
T" r f r^ i Japan
-ired Japan
1 _ist or f annings .
Per cent.
5.84
.-58
Commercial.
Formosa oolong —
Fooohow oolong
Congou.
India
Ping Suey
Country green
Pan-fired Japan
Basket-fired Japan.
Japan siftings
1..30
5. oi
5.31
5.79
5.00
5.20
-5.68
6.71
5.41
6.04
7.12
5.40
5.^
3.59
Per
cent.
6.50
6.49
8.68
6. .50
6.18
7.45
6.02
6.26
6. i<
7.16
6. 48
6.52
7.84
6.04
4.54
6.74
7.00
Per cent.
7.68
7.52
9.25
9.26
7.75
6.73
7.75
6.72
6.70
7.57
8.3e
r.39
r.5S
J.47
r.43
r.6i
5.09
7.61
7.73
^ For description of hvdroeen drying oven used see Ann. Kept. Comi. Agx. Exp.
Sta., 1889, p. 1S7.
51
EEPOET ON rOOD PRESEEYATIVES.
By W. L. Dubois, Associate Referee.
During the past year the referee has confined his attention almost entirely to the
quantitative estimation of salicylic acid. A method was first worked out for the
estimation of a water solution of salicylic acid and attempts were then made to apply
this method to various food products.
Salicylic Acid in Water.
general method.
For this determination ether was selected as the solvent and experiments made to
prescribe certain details of manipulation. The following points were determined:
(1) Four successive extractions with ether remove all but a trace of salicylic acid.
(2) Washing the combined ether solutions twice with 25 cc of water removes all
the mineral acid and only a trace of the preservative. A smaller amount of water
is not so satisfactory.
(3) The use of more than a small amount of alcohol in dissolving salicylic acid
previous to diluting to volume is inadvisable because the color reaction is inhibited
thereby. Warm water is a more satisfactory solvent than dilute alcohol for this reason.
(See Table 8.)
(4) For the most satisfactory matching of colors it is advisable to have the strength
of the standard and that of the solution to be compared therewith approximately the
same. One milligram of salicylic acid in 50 cc is a very convenient amount. It is
also well to keep the temperature of both the standard and the solution approximately
the same.
(5) For producing the color a 2 per cent solution of ferric alum, boiled for a minute
or two and filtered, is very satisfactory. This solution does not cause cloudiness
when used for the test, and an excess of 1 cc does not influence the delicacy of the
reaction.
Three water solutions of sodium salicylate were sent out to determine the most
convenient quantity of ether to use at each extraction, and to learn whether any loss
of salicylic acid resulted from distilling the ether. From the results obtained (as
shown in Table 1) it does not appear that distilling the ether causes loss of salicylic
acid, nor does the use of the larger quantity of ether insure a more complete extraction.
In the opinion of the referee it is safer to evaporate the ether solution until about 20 cc
remain, allowing this last portion to volatilize spontaneously. With this modification
the method is satisfactory for water solutions of salicylic acid.
52
Table 1. — Detennin<ition of salicylic acid in uater solution.
Salicylic acid recovered.
Analrst.
Sample
No.
Salicylic
acid
added.
50 cc ether.
75 cc ether,
50 cc ether,
75 cc ether.
at each
at each
at each
at each
extraction.
extraction.
extraction.
extraction.
evaporated
evaporated
distiUed—
distilled—
spontane-
spontane-
ouslv—
ouslv —
B.H. Smith.
"\V. 5. Richardson.
F. O. Woodruff.
C. W. Harrison.
W. L. Dubois.
mgs.
mgs.
8.0
25.0
45.0
6.0
25.0
50.0
9.2
28.0
46.3
5.6
24.0
45.0
10.0
29.1
50.0
mgs.
8.5
24.0
45.0
4.5
25.0
50.0
8.4
28.5
43.8
5.25
mgs.
48.0
9.43
30.50
50.0
9.0
26.0
48.0
8.0 ;
25.0
50.0 :
8.0 ;
28.5 I
41.3 i
4.6 1
24.0 j
45.0 :
9.3 i
30.5 1
50.0 i.
48.0
10.0
25.0
50.0
9.4
21.0
41.3
47.5
9.3
Salicylic Acid ix Beer.
In applying this method to light beer the referee has found that if at each extraction
a volume of ether at least equal to that of the beer extracted be used little trouble with
emulsions is experienced and the determination is quite satisfactory. A small amount
of coloring matter is found in the ether extract, but the quantity is not sufficient to
cause any trouble. If after distilling all but the last 20 cc the beaker be allowed to
stand oyer night or for a number of hours in a vacuum desiccator, the salicylic acid
will usually crystallize and may be taken up directly in water. The following results
were obtained by the referee:
Table 2. — Determination of salicylic acid in beer.
Sample.
r^-I^^^'tl -imount Percent ^.^^, .^dT,?^o Amount Percent '
m^ found. recovered. ^^°^Ple. ^f^ed^^o ^^^^^ recovered.
A
B
C
D
mgs. mgs. mgs. mgs. ',
10 8.0 80.0 E .30 25.0 , Sa 3
25 22.4 90.0 F 40 .36.4 , 90.9
10 ?.7 87.0 G 50 40.0 80.0
20 17. 9 S9. 3
Salicylic Acid ix Wixe.
MODiriCATIOX OF GEXERAL METHOD.
When applied to wine, either white or red. this method does not give results that
are at all satisfactoiy. The amount of interfering ether extractive is such as to pro-
hibit taking up directly in water. To avoid this difficulty, a common practice here-
tofore has been to exhaust the dried ether extract with low boiling-point gasolene
(boiling point 40°-50° C.>.but we have found that in many cases this solvent extracts
only a small per cent of the salicylic acid present.
In Table 3 are given a few results obtained by extracting four times with ether and
exhausting the dried extract with a low boiling-point gasolene.
53
Table 3. — Determination of salicylic acid in wine.
Analyst.
Amounts of salicylic acid recovered.
16389.
16390.
16391.
16392.
II. V. Frost
mgs.
0.000
.075
.027
.031
Trace.
Trace.
Trace.
.500
5.000
mgs.
0.000
.150
.0536
.014
.133
Trace.
Trace.
2.000
20.000
mgs.
0.000
.100
.0803
.031
.025
Trace.
Trace.
.700
7.500
mgs.
0.000
F 0 Woodruff
.200
.105
A. W. Ogden
.016
B H Smith
.060
C. W. Harrison
Trace.
C. N. Berkeley
Trace.
W. L. Dubois
1.500
Amount of salicylic acid added per 100 cc
17.500
The difficulty in the determination of salicylic acid in wine is due to ttiose substances
which are extracted by ether and interfere with the color reaction. A number of possi-
1)1 e ways of avoiding this trouble have occurred to the referee:
(1) Find a solvent for salicylic acid which will dissolve it from the dried extract and
leave behind other substances.
(2) Sublime salicylic acid from the dried ether extract.
(3) Precipitate out of solution, before extraction with ether, those bodies which
interfere with the color reaction.
(4) Separate salicylic acid from wine with a solvent which will not extract interfer-
ing substances or by distillation with steam. (Not subjected to experiment.)
(5) Precipitate the salicylic acid from the solution instead of extracting. (Not sub-
jected to experiment.)
1. VARIOUS SOLVENTS.
For taking up salicylic acid from the extract gasolene has been proved useless for
quantitative work. In trying several other solvents 100 cc of wine were extracted with
ether and salicylic acid added to the extract. The extract was then rubbed up with
ten 5-cc portions of the solvents in question. The following results were obtained.
Table 4. — Determination of salicylic acid in wine using different solvents.
Solvent.
Salicylic Salicylic
acid ! acid
present, recovered.
Per cent
recovered.
Benzol
Carbon bisulphid
Carboft tetrachlorid. . .
Gasolene (90 per cent)
Ether (10 per cent) a . .
mgs.
mgs.
4
9.5
3.5
5.2
n Larger percentages of ether seem to take up some coloring matter.
Wines containing known amounts of salicylic acid were extracted in the usual way
with ether and the residue treated with ten 5-cc portions of carbon bisulphid. The
results were as follows:
Table 5. — Determination of salicylic acid in wine with ether extraction and carbon
bisulphid.
Wine.
Amount
Amount
Per cent
Wine.
Amount
Amount
Per cent
added.
recovered.
recovered.
added.
recovered.
recovered.
mgs.
mgs.
mgs.
mgs.
White
10
6.07
60.7
Red
10
7.9
79.0
Do
10
7.89
78.9
Do
20
15.0
75.0
Do
20
15.0
75.0
Do
20
17.64
88.2
Do
20
15.0
75.0
54
Chloroform takes up too much, coloring maTier from the residue to be used for this
purpose. For extracting the preservative irom -wine no satisfactory solvent was found
which would not also dissolve interfering substances. Chloroform can be manipu-
lated so as to extract but little tannin, but the referee lias not been able to recover more
than 50 per cent by four extractions with this solvent. Amyl alcohol produces such
heavy emulsions that its use is not practi<a.ble. Mixtures of gasolene and ether in
various proportions did not recover over 40 per cent in four extractions, nor were the
extracts obtained free from coloring matter. ^
2. STIBUMAHOX.
For subliming salicylic add the referee tried the apparatus described by Bigelow
(Btil. 90, p. 59), and also one made by substituting in the abctve a Trolius's nitrogen
bulb for tube (b). It was found that the most satisfactory temperature for completely
subliming salicylic acid is 185° to 195° C. Air should be drawn through at about three
bubbles per ^joond. Salicylic acid was di^olved from^ the collecting tube by dilute
alcohol, diluted to a definite volume and an aliquot portion compared with a standard
solution of salicylic acid. Under these conditions 90 to 100 per cent can be recovered
when working with pure salicylic acid. When heating the extract from wine, how-
ever, an oily, colored substance frequentiy sublimes with the preservative and inter-
feres with the color reaction. The referee obtained but indifferent results, sometimes
recovering 80 per cent, but more frequentiy from 30 to 50 per cent.
3. TEE ATMEXT OF WTSTE BEFOSE EXTRA.CTIOX.
One hundred cubic centimeters of red wine containing 10 mg of salicylic acid we^-e
stirred up thoroughly with 2 grams of washed hide powder and filtered. The filtrate
was extracted in the usual way with ether and 52 per cent of salicylic acid was recov-
ered. One hundred cubic centimeters of wine containing 10 mg of salicylic acid were
treated with gelatin solution and filtered. The filtrate was extracted in the usual way
and 50 per cent was recovered.
One hundred cubic centimeters of red wine containing 10 mg of salicylic acid were
made ammoniacal, ferric chlorid added, and the wine evaporated to 30 cc. The resi-
due was transferred to a graduated flask, diluted to volume, and an sdiquot portion
filtered off. From this filtrate 4.8 mg of salicylic acid were recovered by extractic-n
with etlier.
Fifty cubic centimeters of red wine were saturated with zinc sulphate. The wine
was not clarified by this process.
One hundred cubic centimeters of wine containing 10 mg of salicylic acid were
treated with anitnal charcoal and filtered. Clarification was accomplished by this
ni'^tns. but the salicylic acid is also practically all removed.
Harry wid MumnMrfs method.^ — This method depends on the fact that lead tannate
is insoluble in. caustic alkali, while lead salicylate is soluble. In following the pro-
cedure for wine as laid down by these authors, the referee obtained the unsatisfact^y
results shown in Table 6. By experiment it was found that 5 cc of lead acetate were
sufficient to precipitate all the tannin and coloring matt^ from 50 cc of wine. It was
also found more satisfactory to add hydrochloric acid to the solution before diluting
to volume and filtering, employing the method prescribed by these analysts for deter-
mining salicylic acid in jams, etc. Tlie method as modified by the referee is as
follows:
To 50 cc of wine, add 5 cc basic lead acetate, 30 cc of normal sodium hydroxid,
and partially neutralize with 20 cc of normal hydrochloric acid. Dilute to 300 cc and
filter ^X) cc. Acidify with dilute hydrochloric acid and extract as directed for ]>eer.
The extract usually has some acetic acid in it, and this must be allowed to volatilize
a Analyst, 1905, SO: 124.
55
before taking up the salicylic acid in water. Solution and comparison with standard
is carried out as with beer. From red wines about 75 per cent of salicylic acid may
be re'covered. White wines, which would seem more easily handled, give only about
55 per cent, which fact remains unexplained.
Table 6. — Salicylic acid in wine determined by Harry and Mummery's method.
Wine.
Amount
added.
Amount
recovered.
Per cent
recovered.
Wine.
Amount
added.
Amount
recovered.
Per cent
recovered.
White
mgs.
10
10
10
mgs.
5.5
5.5
7.5
55
55
75
Red
mgs.
10
10
10
mgs.
6.6
7.5
6.8
66
Do
Red
Do
Do
75
68
Salicylic Acid in Tomatoes, Catsups, etc
With this class of goods the material for analysis must receive some preliminary
treatment to remove interfering extractives. In all the methods tried 50 grams
pulped tomatoes were taken and the salicylic acid added.
method 1.
Fifty grams of tomatoes were shaken 30 minutes with 150 cc of water made alkaline
with sodium hydroxid. The mixture was centrifuged, the supernatant liquid poured
through a filter, and an aliquot portion extracted with ether after acidifying. The
residue remaining after evaporating the ether contained considerable coloring matter
and other foreign substances and the method was abandoned.
METHOD 2.
Fifty grams of tomatoes and 100 cc of water were acidified with phosphoric acid
and distilled with steam till 250 cc had passed over. The distillate was made alkaline,
concentrated to 100 cc, acidified, and extracted with ether. No coloring matter or
other foreign substances were present in the ether residue, but neither was salicylic
acid in any quantity. No test was obtained with samples containing up to 100 mg per
kilogram. In one having present 200 mg per kilogram, 0.5 mg was found, correspond-
ing to 10 mg per kilogram.
For such samples, separation of salicylic acid by the above method is not quantita-
tive. A very large volume of distillate is required to carry over any amount of salicylic
acid, and that amount is only a small percentage of the preservative present. The
method is not to be recommended if a better procedure be available.
'^ METHOD 3.
Transfer 50 grams of pulped tomatoes to a 200 cc flask with 50 cc of water, and make
alkaline with milk of lime. Complete to volume and filter as large an aliquot portion
as possible. Usually 150 to 160 cc of filtrate may be obtained. Acidify with dilute
hydrochloric acid and extract with ether four times, using from 75 to 100 cc of ether at
each extraction. Wash the combined ether solution twice with 25 cc of water, and
distil the ether slowly, allowing the last 20 to 25 cc to evaporate spontaneously. Take
up the ethcn* in dilute alcohol, make to a definite volume, using a few drops of a 2 per
cent solution of ferric alum to produce the color. The results obtained by this method
are shown in the following table:
56
Table 7. — Determination of salicylic acid in tomatoes hy Method 3.
ac.TrTr,ia Amount Amount , Per cent | q„~t,i„ Amount I Amount Percent
oampie. j ^dded. found. ' recovered. *a™P^e. added. I iound. recovered.
mgs. I
2.5 !
5.0
10.0
0.0
1.82
5.0
0.0
36.4
50.0
-gs.
mgs.
20.0
12.5
25.0
16.92
30.0
22.70
In the ether extract from sample A. crystals appearing to be salicylic acid were pres
ent, but no test for salicylic acid was obtained. This led to aji investigation of the
effect of alcohol on the color produced by ferric salts and salicylic acid and the follow-
ing experiment was made:
A solution of 1 mg of salicylic acid in 50 cc of water, to which were added 3 drops of
ferric solution, was used as a standard. The solutions matched against this contained
1 mg of salicylic acid and various quantities of alcohol in 50 cc. as shown in the table
below:
Table 8. — Effect of alcohol on color produced hy ferric salts and salicylic acid.
■Readm? ' ^^eading Salicylic
-Ucohol fit^rirf ofsolu- acid in-
mined, solution.
Remarks
20 1 '
18 1
18 ■• 1 Quality of color not the same.
21 Color of sample decidedly bluer.
27 I Color qualltv identical.
37 I ' Do.
Color of sample too light to read.
Do.
It appears from these results that the presence of more than 10 cc of alcohol in the
solution used is inadvisable. It is preferable to dissolve the ether extract in warm
water, cool, and dilute to volume.
It is also an improvement to render the tomatoes alkaline with ammonia before
adding the milk of lime. When this is done, about 15 cc milk of lime (200 grams of
quicklime in 2.000 cc of water) are sufficient, whereas much more is necessary when
the ammonia is not used. These two modifications in the method given ab >ve have
solved the problem and give excellent results, as is shown by the following figures:
Table"9. — Determination of salicylic acid in tomatoes hy modified jnethod . {Dubois.)
Amount
Amount
Per cent
Amount
Amount
Percent i
used.
recovered.
recovered.
used.
recovered.
recovered.
mgs.
i mgs.
1
mgs.
jjig.s.
0
47
94.0 !
20
19.20
96.0
10
ao
80.0
25
25.00
100.0
10
8.11
8L1 i
30
26.70
1 89.0
15
13.33
88.8 1
! 1
50
46.9
1 93.8
57
Two samples were sent out to the collaborators containing, respectively, 10 mg and
20 mg of salicylic acid in 50 grams of tomatoes, to be examined by the above method.
The results obtained on these samples are given in the table below:
Table 10. — Results obtained by collaborators on salicylic add in tomatoes by modified
method.
Analyst.
No. 17073
(10 mg
added).
No. 17074
(20 mg
added) .
C S Brinton
7.44
7.50
4.53
3.0
17.05
G. F. Mason
17.50
R . Hoagland
14.86
B. H. Smith
17.50
Salicylic Acid in Jams, Marmalades, and Similar Products.
The referee has found that Harry and Mummery's method is fairly satisfactory for
this class of products, recovering from 60 to 75 per cent of the salicylic acid added.
The method is as follows:
Fifty grams of the crushed sample to be tested are placed in a 300 cc flask with a
small amount of water. Ten to 20 cc of a saturated solution of basic lead acetate are
added and the whole made alkaline by the addition of 25 cc of roughly normal sodium
hydroxid. Shake well and add from 15 to 20 cc of roughly normal hydrochloric acid.
The contents of the flask are then diluted to the mark, well shaken, and an aliquot of
200 cc filtered off. This is acidified with hydrochloric acid and extracted with ether.
No comparative work has been done with other methods. In the table below are
given a few results obtained by the referee:
Table 11. — Determination of salicylic acid in jams, etc., by Harry and Mummery^ s
method. (Dubois.)
Amount
added.
Amount
recovered.
Per cent
recovered.
Amount
added.
Amount
recovered.
Per cent
recovered.
mgs.
10
10
20
mgs.
6.42
6.17
15.00
64.2
6L7
75.0
mgs.
30
10
25
mgs.
15.00
6.17
14.06
50.0
6L7
56.3
Recommendations.
The referee recommends the following changes in Bulletin No. 65:
(1) Page 109, that the second method for the detection of benzoic acid be rewritten
to read as follows:
Evaporate to dryness and treat the residue with 2 or 3 cc of concentrated sulphuric
acid. Heat until white fumes appear. ^The organic matter is charred and benzoic
acid is converted into sulpho-benzoic acid. A few crystals of ammonium or potassium
nitrate are then added and the dish again heated until white fumes appear. Repeat
this process until all organic matter is oxidized and the solution is practically color-
less. This causes the formation of metadinitrobenzoic acid. When cool, the acid
is diluted with an equal volume of water, ammonia added in excess, and the solu-
tion transferred to a test tube. This is cooled and a few drops of freshly prepared
ammonium sulphid added in such a way that the solutions do not mix. The nitro
compound becomes converted into ammonium metadinitrobenzoate which possesses
a red color. This reaction takes place immediately and is seen at the surface of the
liquid without stirring.
(2), Page 90, line 2, substitute "20 cc" for "5 cc." This change is considered
advisable because when determining sulphurous acids in meat products about 20 cc of
20 per cent phosphoric acid is found to be necessary to completely evolve the sul-
phurous acid present. This amount would not be required for wines, but, as there
58
seems to be no objection to the use of the excess of phosphoric acid in this case, it is
thought advisable to recommend a uniform quantity of phosphoric acid which will
answer for all purposes.
(3) Page 90. after line 34. insert •"Instead of titrating the excess of iodin with the
standardized thiosulphat^ solution, the sulphui'ous acid distilled over may be deter-
mined directly by acidin-ing the iodin solution or an aliquot thereof with hydrochloric
acid, boiling until colorless and precipitating sulphuric acid in the usual way with
barium chlorid . ' "
(4) Page 110. line 41, transpose the phrase ••boiled to expel carbon dioxid "' to line 40,
inserting the same after the word "pink.""
(5) Under method for boric acid insert the following method:
Weigh about 50 grams, make alkaline with milk of lime, evaporate to dr^mess and
ash. Dissolve in hydrochloric acid, filter and wash. In case much carbon remains,
bm-n the paper and residue after making alkaline with milk of lime and treat with
hydrochloric acid as above. Make the tiltrates alkaline with sodium hychoxid. boil,
add barium hydroxid tmtil no fm-ther precipitate is formed, and filter. Dissolve the
precipitate in dilute hydrochloric acid and reprecipitate with sodium hydroxid with
the addition of a few drops of bariiun hydroxid. Wash xhe precipitate with hot water,
cool the filtrates and washings to room temperatiue. and dilute to definite volume.
Take an aliquot portion, add methyl orange, and acidify-. Boil to expel carbon dioxid,
cool and add decinormal alkali imtil the pink color is just discharged. Add 5 or 6
grams of mannite and a few ch'ops of phenolphthalein and titrate the boric acid, which
is now all in the fi-ee state, with decinormal alkali. When the end point is obtained
it is well to add more mannite to be sure that enough has been used.
(6) Eliminate irom. methods for salicylic acid directions to extract the residue with
gasolene.
(7) Page 73, eliminate the resorcin method for formaldehyde, inasmuch as this
method is unreliable and there are better ones available.
It is further recommended that the referee for the ensuing year conduct work along
the following lines:
(1) Quantitative determination of salicylic acid in wines.
(2) Further experiments for the determination of salicylic acid in fi'uit products
such as jams, marmalades, etc.
(3) Further trial of Harry and Mummery's method.
(4) Trial of method for the detection and determination of fiuorids proposed by
Woodman and Talbot iJ. Amer. Chem. Soc, 1906, :2S: 1437 .
The privileges of the Cosmos Chib were extended to the convention
through the secretary of the association, and the meeting adjourned
initil 2 oVdock.
WEDNESDAY— AFTERNOON SESSION.
EEPOET ON DETEEMIi^ATIOlT OP WATEE IN POODS.
By F. C. Weber. Associate Referee.
The study of this subject by the association was authorized at the 1905 meeting.
when a resolution was adopted instructing the referee on food adulteration to provide
for the determination of moistm'e. stud^"ing particularly Benedict's vacuum method. a
On August 6. 1906. the following circular letter was sent to eleven chemists who had
previously signified then interest in this subject, asking then* cooperation for the
present year:
Deab Sir: At the last meeting of the Association of Official Agricultural Chemists,
held at Washington. D. C. November 16-18. 1905. the referee on food adulteration was
a Benedict and Manning. Amer. J. Phvsiol.. 1905. 13: 309.
59
instructed to provide for the determination of moisture, studying Benedict's vacuum
method and Maquenne's method.
Please inform me at your earliest convenience if you can assist in this work, so that
samples and instructions can be forwarded as soon as possible.
There will be four samples sent out, the moisture to be determined by the methods
above and by the method in use in your laboratory.
Respectfully, F. C. Weber,
Associate Referee on Food Adulteration.
Replies were received from eight chemists, all but two stating that pressure of other
duties prevented them from taking up the work at this time, and the two who replied
favorably did not send reports. Three collaborators were secured in the Bureau of
Chemistry.
The samples selected this year were:
No. 1. Finely bolted rice flour.
No. 2. Durum wheat flour.
No. 3. Potato starch, prepared by grinding in a porcelain mortar.
No. 4. Shredded wheat biscuit, prepared by grinding in a burr mill.
The first three samples were in a fine state of division, while the sample of shredded
wheat- was somewhat coarser. Each sample was thoroughly mixed and allowed to
stand at room temperature for twelve hours before being bottled and sealed with
paraffin.
Preparation of the Vacuum.
Benedict in his original method « employed the Hemple form of desiccator to obtain
a high vacuum by chemical means. One hundred and fifty to 200 cc of concentrated
sulphuric acid are placed in the upper compartment of the desiccator and 10 to 20 cc of
anhydrous ether, allowed to flow from a pipette, are placed in the bottom of the lower
part. After adjusting the top suction is applied, and when the pressure is diminished
sufhciently the ether begins to boil, the vapors forcing out the air within the desiccator.
When all the air is removed by the suction, the stopcock is closed and the remaining
ether vapor is absorbed by the sulphuric acid.
A vacuum of 1 to 4 mm can be obtained by this method in fifteen to twenty minutes.
It is essential to have a manometer within the desiccator. It was found advisable to
place between the desiccator and the exhaust two wash bottles, the one near the desic-
cator acting as a trap, while the other contains water through which the ether bubbles,
showing the rate at which the system is exhausted. When the water just begins to
start back the stopcock on the desiccator is closed.
Recently it was found by H. C. Gore & that the Scheibler form of desiccator works
equally as well as the Hemple type for the drying of substances in a vacuum. There is
no essential difference in the use of these two forms, the sulphuric acid being above
the samples in the Hemple desiccator, while in the Scheibler form it is below them.
The vacuum is obtained in the same manner, the ether being placed in some convenient
receptacle which floats or stands in the sulphuric acid. Inverted ground-glass stoppers,
tall enough to be just above the surface of the acid, were found to be quite convenient.
Both t^'pes of desiccators were used in this work. To avoid errors in weighing samples
of this nature in open air, small aluminum dishes, 4.5 cm in diameter and 1.5 cc in
depth, provided with a tightly fitting cap of the same material, were employed.
It was hoped that a number of collaborators would respond, that a comparison might
be made of the various methods employed in different laboratories for determining
water in agricultural products. Simplie as this determination may seem, it is one of
the chief sources of error in the calculation of analytical results.
a Benedict and Manning, Amer. Chem. .!., 1902, 2: 340.
bj. Amer. Chem. Soc, 1906, 28: 834.
60
In the third volume of Wiley's Principles and Practice of Agricultural Analysis, 23
pages are devoted to descriptions of methods and apparatus for drying organic bodies.
Although the principles involved in each division and subdivision of the methods
are practically the same, the application of each method to the same sample would
probably give a series of different results, the variations in the details of each method
having a marked influence.
Carr and Sanborn « in 1895, in an exhaustive study on the desiccation of organic
liquids, called the attention of the association to the variation in moisture results.
They showed that this variation was caused solely by the decomposition and oxidation
of the sample and that it was a function of the temperature to which the sample was
exposed. By employing a vacuum ovenb through which a current of dry air
passed, they were able to obtain constant results on sugar solutions (levulose) which
remained constant after four hours drying.
Determinations were made in this work under the following conditions:
1. In a partial vacuum (25^^-28^^) at a temperature of 100° C, through which a slow
current of air, dried by passing through sulphuric acid, flowed.
2. In a current of dry hydrogen.
3. In the Hemple and Scheibler vacuum desiccators.
Approximately 1 gram samples were used. The nature of this set of samples is such
that the first method very likely gives the maximum results in the shortest time of
drying. On inspection of the tables this is seen to be the case, and the results of this
method are therefore used as a basis for a comparison of the results obtained by the
other methods.
Weighings were made according to the vacuum method at the end of 24 hours, 48
hours, 3 days, 5 days, 7 days, 12 days, and 20 days. In the determinations made in
hydrogen, weighings were made at the end of each hour, the maximum time of heating
being from 6 to 8 hours.
Table 1. — Moisture determinations in duplicate ohtained hy drying in air for varying
periods.
J. S. Chamberlain.
H. C.
Gore.
S. Leavitt.
F.
C. Weber.
Samples.
6
hours.
11
hours.
15
hours.
1
hour.
2
hours.
4
hours.
10
hours.
5
hours.
10
hours.
15
hours.
1. Rice flour
Per ct.
i 11.60
1 11.63
Per ct.
11.62
11.67
Per ct.
11.62
11.67
Per ct.
11.67
11.60
Per ct.
11.70
11.62
Per ct.
11.26
11.22
Per ct.
11.46
11.44
Per ct.
11.18
11.24
Per ct.
11.40
11.38
Perct.
11.42
11.46
[ll.62
.11.64
11.64
11.64
11.66
11.24
11.45
11.21
11.39
11.44
2. Durum wheat
flour
f 11.33
1 11.35
11.31
11.39
11.32
11.38
11.18
11.16
11.27
11.27
11.06
10.98
11.20
11.18
11.04
11.03
11.20
11.22
11.23
11.24
ill. 34
11.35
11.35
11.17
11.27
11.02
11.19
11.04
11.21
11.24
3. Potato starch....
( 14.59
1 14.61
14.56
14.63
14.58
14.63
14.64
14.65
14.60
14.69
14.24
14.14
14.38
14.34
14.33
14.27
14.50
14.46
14.49
14.43
[l4.60
f y.24
1 9.21
14.60
14.61
14.65
14.65
14.19
9.04
8.94
14.36
9.24
9.20
14.30
8.99
9.05
14.48
9.17
9.20
14.46
4. Shredded wheat .
9.28
9.24
9.31
9.25
8.78
8.86
9.08
9.10
9.17
9.20
I 9.23
9.26
9.28
8.82
9.09
8.99
9..2
9.02
9.19
9.19
In Table 1 are given the results of the determinations made in the vacuum oven,
through which a current of dry air passed. The temperature of the oven was kept at
«U. S. Dept. Agr., Bureau of Chemistry, Bui. 47, p. 134, "The dehydration of
viscous organic liquids. ' '
b Designed by Carr; described in Wiley's Principles and Practice of Agricultural
Analysis, 3: 23.
61
100° C. and the vacciim at about 25 inches, which corresponds to a water-l)oiling point
of 57°. Thus the water in the samples was exposed to a temperature 43° abov(^ the
boiling point.
Considerable variation is seen in the time of drying employed by different analysts
even in this laboratory. Since the kind of material has the greatest influence on the
time of drying, this point should be regulated by future work and not be left to the dis-
cretion of the analyst. This is seen in sample No. 1, which contains a high percentage
of starch, and No. 3, which is pure potato starch, in which cases drying for 2 hours gave
practically the same results as drying for 15 hours, while in samples Nos. 2 and 4 there
is a smaller percentage of water obtained by drying for a shorter period of time. The
relative fineness of the samples also exerts its influence on this point.
The determinations in the last two instances in this table were made at a later date
than the first two and after the samples had accidentally stood a short time, in the
laboratory, unsealed.
62
6
"A
i
1
1
i
i
1
i
1}
1
li 1
]
:
I
6
c
"5.
£
■
ii
1
ill
V?
X
?.
?
X ■ «X 5t
t- Xi>- -r
ci ©St d
i i
-^2S
|q6qo
00
E3§
00 00
00
ON
1-1
e3
o
.-1
o
2 '
2
!
!
§2
00 00
1
OS
1
8
d
©
a
^8 i
32 ©
r
1
d
d
5
©
ldd
©
x2
^1
•S 2
©
©
o
£5
d
-2
god
1
■
ii 1
CO
i c
; ^
, o
S
i
1 .
1!
1
*«
«
:
S3
?
o
00
s|§s
?
§
"
•*
; C
' 2,
^^1 ill
©
§s
1-*
{2
d
i
©
« II
vL
Si- go!d
1-
Oj4oO
©
1
1
i
.ill:
'w
1-;
§2
Ii 11
-^ — V
2
^
Dried
24 hours.
5^
©
22§:::|
22s5
©
^2
1
•
1
1
i
1
1
1
1
li
i
1
1
1
i .
1
-<
1 :
r <
: >
: <
1
i
<
i
s
<
i
I
1
Zf
s
>
<
63
^
§
f
~
i
6. 77 6. 76
6.68 6.57
6.62 7.20
6.67 7.07
^
W
1
«
-
-
-
CO
^
■*
•^
■*
^
^
OS
1
^
00
^
•*
■*
^
^
*
■*
n
«
^
t.
<
?
J
>
3
.CO
CO
X
V.^
©
I
«l
1
1
!
. '~-,
,_^
^-^
P
kCJ-
o
00
©
^'^
^^
Q^
to o CO t^ cc
^?
eo^t^coco
X®
■f^^
Oi->
P^
<u
' l-* ^
cl
fts
.00 -co
<N
■«o tT
t^
^
-g"^
M "^
is
o
©
&
n'^
<u
OQ
as
'^^
St^t-O5C0
J2«e
■f^iS
©o
Q'^
a^
. t^
TT
«
li
d
©
fiu,
a^
^^
^
"E 2
^y
©
ot:,
<ii
a.
ii
^S
^
^
p2
'"'
as
. /—v
,,-v
,..^
Is
^
i>
<u
as
Ml>-
^-j-Hj-.^
^r^
P^
<»
' i-<i^
rf
as
^^
^
<N
^
'g>.
<D
u =3
CJ 1—1
■—1
,^
a
P'^
^
S
a.
"^^
lU . . .
S^
■C.5
tH 1—1
P'*^
' 1-1^
as
-..2
^^
S
■*
•C o
""1
^
«.:
;..
^
as
. T><
CO
o
TiiS
^"!
©
■E o
"—I
,^
1
P-^
^
U
.
ci'
a.
-— v-^
i
>.
"3
c
,.,
<
03
M
g
01
a>
>
^
<J
O
f^
1
|l
Soc
00
05
00
«1
. .— N
..-^
,~^
i
'^^^
<*
SO
1
X
p;^
fl^
•»5.-Ht^ot^
SSOS
^ 00 00 00 oc
©-*
. XX
Q'^
d
an
^S3
^
??
^
'g^>;
sjt-^
t-
■n OS
a
p^
1
1
as
^-N^
II
eS3^SEH
Xt>-
wt^t^t^t^
^ t-r-
p-^
%>
~-'
a,
00
•ki CO
a>
12
ffl 3
St-
t-
J^
•m o
«f
t
as
^2
.a
n
•^§
■^
Qi
i
^
a^
p^
11
Tf5
■>*l
as
v.^
X
■*
p;^
s^d-
■—I
■u
as
/-v^
V ^-v
-g?
t-X
■^^
y Tti Tfi rr ■*
-t*-*
p*^
u
' '" ^*
CO
6
^-
as
^s
OS
^
-g^
■^ ^
^ -^
p<
p-^
V
S
as
v.^
goSo?
gg
"CJ
cj Ti5 -q? TlJ Tt
•-ft
p'*^
^
as
in
^•^
i;^
^
« 3
^^
^
■*
P-^
K
%>
1
4i
as
.(M
t-~
i>
-rt2
so
•Co
^ -^
3
■*
P-«
!^
(M
(^
_^
"3
a> 1
<!t
1
>
(*
i
fe «
= a
64
In Table 2 the results obtained by drying in the Hemple and Scheibler desiccators
are given. The results obtained after standing 7 days in the Hemple desiccator
compare very favorably with one another, though in a few instances maximum results
are obtained even with 5 days' drying. The maximum results obtained for 7 days'
drying are 11.56 per cent for sample No. 1; 10.86 per cent for No. 2; 14.73 per cent
for No. 3; and 8.30 per cent for No. 4. The average results are, for No. 1, 11.28 per
cent; No. 2, 10.49 per cent; No. 3, 14.50 per cent; and No. 4, 7.90 per cent.
In that part of the table showing the results obtained in the Scheibler type of desic-
cator the maximum figures show uniformly higher results, being nearly 0.2 per cent
higher than the maximum figures obtained in the Hemple form. Further, as the
determinations were carried on in this case for periods of 12 and 20 days, the results
indicate that drying for 7 days extracts the greatest amount of moisture from the
samples. In sample No. 1 there is a gain in weight of the sample after 7 days' drying,
No. 2 and No. 3 remain practically constant, while sample No. 4 shows a gain.
Table 3. — Moisture determinations obtained by drying at a temperature of boiling water
in a slow current of dry hydrogen.
No. of Dried
sample., 1 hour.
Dried
2 hours.
Dried - Dried Dried
3 hours. : 4 hours. 6 hours.
Dried
8 hours. 1
Per cent.
1 9. 98
2 11. 17
3 13. 63
4 1 8.73
Per cent. Per cent. \ Per cent. Per cent.
10.99 11.13 11.24 11.53
11.33 1 11.43 11.46 1 11.56
14.51 14.48 14.51 ! Gained.
9.04 9.10 9.11 1 9.21
1 1
Per cent.
11.79
Gained.
Gained.
Table 3 gives the results obtained by drying in a current of dry hydrogen at the
temperature of boiling water. Determinations were made in duplicate and triplicate,
the average only being given. Weighings were made at the end of each hour's drying
and continued until the sample gained in weight. The maximum time of drying
was 8 hours in the case of sample No. 1. while 4 hours' drying sufficed for sample No. 3.
Table 4. — Comparison of maximum moisture determinations given by each method.
No. of
sample.
Partial
vacuum-
current
dry air.
Hemple
vacuum
desiccator.
Scheibler ; Current
vacuum ; dry
desiccator, hydrogen.
1"
2
3
4
Per cent.
1L66
1L35
14.65
9.28
Per cent.
n.56
10.86
14.73
8.30
Per cent.
IL 79
a 11. 08
14.89
8.49
Per cent.
1L79
n.56
14 51
9.21
a 20 davs in a vacuum.
In Table 4 the maximum results obtained by each method are arranged for com-
parison. In the samples containing the largest amount of starch, Nos. 1 and 3, the
vacuum method gives the highest results, while in the more complex samples, as No. 2,
wheat flour, and No. 4, shredded wheat, drying in partial vacuum in a current of dry
air gives the highest results. It is not to be inferred from this that the method giving
the highest results necessarily gives the exact percentage of hydroscopic moisture.
^Yhether material other than water volatilized at the temperatiu'e of 100° C, to which
the samples in partial vacuum were exposed, must be determined by a future study
of this subject.
It is evident that in a vacuum of 0-5 mm of mercury, which is maintained in the
desiccators, there would always be a tendency to establish an equilibrium, the rapidity
with which it would be established depending on the volatility of the substances.
Under such conditions, there must be a small quantity of vapors of sulphm'ic acid at
65
all times within the desiccator, which may exert a slight influence on the determi-
nations.
The fact that determinations made in the Scheibler vacuum desiccator give maxi-
mum results at the end of 7 days' drying and then show a gain in weight in nearly
all cases after that period, led to the determination of total sulphur in the original
samples and in the samples which had stood 12 and 20 days in the desiccator. The
sodium peroxid method was employed, the fusion being conducted in nickel crucibles.
Three-gram samples were employed in duplicate for the determinations in the original
samples, while the vacuum samples were approximately 1 gram. The duplicate
determinations agreed closely throughout and a blank on the maximum quantity of
sodium peroxid used to complete fusion did not give any sulphur.
The results are given in Table 5, and with the exception of No. 3 the figures are quite
striking, showing that this is a factor to be given some consideration, particularly in
reference to time of drying. Sample No. 3, which, it will be remembered, is nearly
pure starch, gave the highest percentage of water by this method and evidently did
not, during the time stated, absorb any of the sulphuric acid. On the other hand,
No. 1, rice flour, also contains a high percentage of starch, but shows quite an increase
in sulphur content. Sample No. 2 shows an increase of 0.048 per cent during the 12
days, while No. 4 shows an increase of 0.098 per cent. The table also shows that
the maximum gain in sulphur takes place within the first 12 days, the results
from standing 20 days being practically the same, and very good duplicates of the
first set.
Table 5. — Sulphur in samples before and after drying in vacuum desiccator.
Description.
No. 1.
No. 2.
No. 3.
No. 4.
Per cent.
0. 103
.141
Per cevt.
0.164
.212
.210
Per cent.
0.019
.017
.024
Per cent.
0.129
.227
In Scheibler desiccator 20 davs... . .
.222
From the results of the present year's work it is evident that the vacuum method
is worthy of careful consideration. The application of the method to a greater variety
of materials will secure valuable data as to its value for the determination of water
in foods. Recommendations with this end in view and the study of any modifica-
tions which may facilitate drying are accordingly made.
Recommendations.
It is recommended that the study of the vacuum method for the determination of
water be continued next year.
That the methods employed for study shall be:
The present vacuum method in either the Hemple or Scheibler type of desiccator.
The vacuum method with any modification which may facilitate drying, as the
introduction of phosphorus pentoxid as a drying agent.
Drying in a partial vacuum in a current of dry preheated air and inert gases.
The report was referred to Committee C on recommendation of
referees.
31104— No. 105—07 5
66
EEPOET ON OEEEAL PEODUOTS.
By A. McGiLL, Associate Referee.
The definitions of cereal products authorized by the Secretary of Agriculture « are
carefully framed so as to employ no words in regard to whose meaning there can be
any doubt. Thus, in the definitions of flom- and meal, no mention is made of gluten
nor of proteids; nor is any minimum of proteid nitrogen fixed. In our present state
of imperfect knowledge this is well, but it is undoubtedly to be desired that such a
well-known and widely used term as gluten could be given a meaning of sufficient
precision to permit of its being employed in connection with flour. It is the word
best known to the miller and the baker as representing the valuable and distinctive
proximate component of wheat flom-. The separation of gluten, the determining of
its total amount and of its qualities, are everyday proceedings in all technical labora-
tories, and especially in such as make a specialty of cereal work.
In the appended notes the recent literature of this subject has been abstracted,
presenting very briefly the novel features in the hope that such a definite conception
of gluten may be reached as will permit of formulating an acceptable definition.
Gluten.
coxditions affectixg gluten estniation.
The term gluten is applied to the residue obtained by kneading a dough from wheat
flom' in a stream of water in such a way as to wash away most of the starch, or until
the wash water remains practically clear. The operation is incapable of yielding exact
results, and when carried out as described can be regarded as affording merely approxi-
mate information about the samples of flom' tested. Arpin^ has shown that the per-
centage of gluten obtained varies with the temperature of the wash water, being higher
with warm than with cold water. Thus, on a sample of wheat floiu*, washed with water
at 5° C, he found 7.83 per cent gluten; at 15° C, he found 8.08 per cent, and with
water at 25° C. he found 9.24 per cent.
Again, if washing be continued after the removal of the starch, a very considerable
loss of gluten occm's. For only five minutes' extra washing c he found a loss of 2J per
cent moist gluten (0.9 per cent dry).
Balland has shown that the amount of gluten obtained depends to some extent upon
the length of time that the cake of dough is allowed to lie before being washed; and
Arpin,^ while coiToborating Balland's observation, shows that. the "hardness" of the
water used in washing has a great influence on the amount of gluten obtained. He
found as much as 4.7 per cent increase (dry gluten) when hard water was employed.
Recognizing the various factors which affect the yield of gluten, Flem'ente advises
the use of water containing 0.1 gram of calcium carbonate per liter (of course dissolved
as bicarbonate) at a temperature of 16° C, kneading the dough for 11 minutes and
washing for 2 minutes, finally drying at 100°-105° C. He has found that distilled
water reduces the yield of gluten; also that lime as sulphate or chlorid gives a
lower yield of gluten than when present as carbonate. Sodium chlorid has a like result.
Fleui'ent also advises that old or acid fiour be made neutral with bicarbonate of soda
before the determination of gluten.
0- Circular 19 of the Secretary's Oflice, 1906, Standards of Purity for Food Products.
& Ann. chim. anal, appl., 7: 325— Abst. J. Soc. Chem. Ind., 1902, 21: 1417.
c Ann. chim. anal, appl., 7; 416— Abst. J. Soc. Chem. Ind., 1903, 22: 168.
(?Loc. cit. 7; 376— Abst. J. Soc. Chem. Ind., 1902, 21: 1560.
e Comptes rendus, 1905, 99— Abst. J. Soc. Chem. Ind., 1905, 24: 155.
67
COMMENT.
The process of crude gluten estimation quite loses its simplicity, and hence its chief
advantage, when all the precautions above described are observed, and it is scarcely to
be wondered at that Arpin and others advise the determination of total nitrogen, and
the use of a factor for converting this into gluten, as simpler and more reliable than
the direct separation of the gluten itself.
CRUDE AND TRUE GLUTEN.
The explanation of the variable results referred to is found in a study of the gluten,
as separated by the usual processes. Norton « has recently given an excellent resume
of the subject in the Journal of the American Chemical Society. Without attempting
to make a full abstract of Norton's paper, it may be said that he distinguishes between
crude gluten and true gluten, the latter consisting, according to the researches of
Osborne and Voorhees& of gliadin and glutenin only. Crude gluten, as usually
obtained, contains, in addition to gliadin and glutenin, variable quantities of fiber,
starch, and other matters. By using a 1 per cent solution of sodium chlorid Cham-
berlainc obtained a gluten containing 13.19 per cent of nitrogen, equivalent to 75.18
per cent of proteids (NX5.7), or true gluten, in the crude gluten. Analysis of a
sample of crude gluten, obtained from a mixture of six durum wheat flours, by the
ordinary process of washing, gave Norton^ the following results:
COMPOSITION OF CRUDE GLUTEN.
Per cent.
Total protein (N X5.7 ) 80. 91
Ether extract 4. 20
Fiber 2. 02
Ash 2. 48
Carbohydrates, other than fiber 9. 44
Total 99. 05
Examination of the proteids, by fractional solution, gave the following results:
Per cent.
Gliadin (70 per cent alcohol extractive) 6 42
Glutenin (0.2 per cent potassium hydroxid extractive) 5. 76
Globulin (10 per cent sodium chlorid extractive) 1. 11
Total proteids recovered 13. 29
Total protein (Nx5.7) as recovered directly from the crude
gluten 13. 39
The results may be written thus:
- Crude gluten.
Per cent.
Fats, or ether extract 4. 20
Carbohydrates (nonfiber) 9. 44
Fiber 2. 02
Ash 2. 48
Gliadin 39. 09
Glutenin 35. 07
oj. Amer. Chem. Soc, 1906, 28: 8.
b Amer. Chem. J., 1893, 15: 392.
cU. S. Dept. Agr., Bureau of Chemistry, Bui. 81, p. 121.
d Loc. cit.
68
Per cent.
True gluten 74. 16
Sodium chlorid. soluble protein 6. 75
Total 99. 05
The total crude gluten (dry) found was 16.55 per cent of the weight of the flour
and the protein found 13.39 per cent.
The content of true gluten corresponds well with that found by Chamberlain, and
Norton concludes that crude gluten, as usually prepared, contains approximately 75
per cent of true gluten.
RELATIOX BETWEEN CRUDE GLUTEN AND TOTAL PROTEIDS.
There is a general agreement between the dry gluten (crude) as found by usual
methods and the total protein (Xx5.7) of flours; but closer examination shows that
there is a tendency for the gluten to exceed the proteids in straight and low grade flours,
to agree closely in patent flours, and to fall short in whole-wheat meal. Norton has
.carefully examined the causes of this difference, and concludes as follows:
"Crude gluten is an expression, in addition to the true gluten content of a flour, of the
balance between the loss of nongiuten proteids and gain from the retention of non-
proteid substances. The relation of the crude gluten content to the total proteid con-
tent can thus be explained by the varying composition of the different flours in respect
to nitrogenous compounds and nonproteids.
He is of opinion that crude gluten is a very rough expression of the gluten content of
a flour or wheat, and the determination has but little worth in the valuation of flours;
that, further, the best simple method for ascertaining amount and character of gluten
is the determination of total nitrogen, with expression of the ratio of gliadin to total
protein.
If we restrict the term gluten to "true gluten "" as above defined, the nongiuten
nitrogen of flour consists of (1) globulin nitrogen, (2) amido nitrogen, and (3) albu-
men and proteose nitrogen. From the results of work by Norton and Snyder it appears
that the following variations occur:
Nongiuten nitrogen of total nitrogen.
Per cent.
Straight durum wheat flour 18. 65
Patent spring wheat flour 14. 97
WTiole durum wheat meal 23. 60
Patent spring wheat flours, mean of two samples 13. 80
Bakers' grade flour 12. 93
VALUATION OF FLOUR.
Kraemer« suggests a classification of flours, as follows:
(I) Those that produce a stiff and cohesive dough in the proportion of 14 to 15
grams of flour to 10 cc of water.
(II) Those that do not produce a stiff and cohesive dough under these conditions.
These are further subdivided into:
(A) Those that form a smooth jelly-like paste upon boiling 1 gram of flour with 15
cc of water for about one minute.
(B) Those in which a more or less granular or liquid paste results.
This subclass may be further subdivided into :
(a) Those which give off an odor of roasting corn when heated in glycerin to boiling
for a few minutes.
a J. Amer. Chem. Soc, 1899, 21: 650.
69
(6) Those not giving such odor.
Guess a states that no limit has yet been reached beyond which the increased gli-
adin content of the gluten is not an improvement in the quality of the flour. In
order at once to express the total gluten percentage and the quality of the gluten as
regards its gliadin content, Guess suggests a composite factor for denoting the grade or
quality of the flour. This composite factor is the per cent of gluten multiplied by
the ratio of gliadin to glutenin. In the application of this factor to a large number
of flours he found a variation from 58.38 to 15.38, and among whole wheats the following:
No. 1, hard— variation from 27.62 to 10.18.
2, hard — variation from 24.58 to 11.97.
1, Northern — variation from 21.11 to 7.20.
2, Northern— variation from 13.28 to 7.33.
Snyder & corroborates the statement of Guess in regard to the importance of the
gliadin content, and holds that the gliadin percentage gives a more valuable datum
than the gliadin glutenin ratio. He emphasizes the difficulty of estimating gliadin
with accuracy, and suggests that gliadin may not itself be a single proteid.
Snyder recognizes the importance of the ash and moisture determinations in fixing
the A^alue of a flour for baking, as also the acidity, but acknowledges that no accurate
methods exist for determining either moisture or acidity in flours.
The baking test is still the most reliable one in ascertaining the value of a flour in
bread making, but the chemical analysis enables nutritive values to be determined.
Color has always been considered as of great value in classifying flours. Snyder c
points out that, in consequence of artificial bleaching processes, this determination
loses rnuch of its value. The blending of bleached low-grade flours with those o£
higher grades is best detected by determining the ash.
Fleurent^ points out that ozonized oxygen has no bleaching effect on flour, and'
peroxid of nitrogen, to the amount of 15 cc to 40 cc per kilo, must be used. The
chemical composition of the flour is not changed, so far as known. The action is con-
fined to the yellowish oil of the wheat, but is not a process of oxidation. It corre-
sponds with a lessening of the iodin value, which changed as follows in three samples:
lodin value.
No. of Before
sample. ! bleaching.
After
bleaching.
■
1
2
3
86.44
86. 10
65.20
81. 70
80.79
56.70
The film of oil becomes transparent and permits the whiteness of the starch to show
through.
Bleaching by age. alone involves oxidation and a precipitation of white, fixed,
fatty acids. Fleurent claims that the enzymes of the flour are not affected by bleach-
ing; that the flour keeps better, and that lower grades of flour are not amenable to the
treatment.
He outlines a process for the detection of bleached flours, based upon fixation of
nitrogen peroxid by the oil. The oil is extracted by petroleum and dissolved in amyl
alcohol and treated with alcoholic potash. An orange-red color results with bleached
flours. The test is capable of showing as little as 5 per cent of bleached flour in
admixture.
a J. Amer. Chera. Soc, 1900, 22: 263.
6 Ibid., 1905, 27: 1068.
cComptes rendus 1906, 180— Abst. J. Soc. Chem. Ind., 1906, 25: 194.
70
Wender« finds tliat wheat and otlier grains contain an enz\Tne capable of liberating
oxygen n-oni hydrogen peroxid and that this enz^Tae resides chiefly in the embryo
and outer coats of the gi-ain. He proposes to use the volume of oxygen liberated in
thii-ty minutes, under fixed conditions, as a measure of the character ''fineness^ of the
floitr. He re-cords the following results for 100 gi-ams of substance:
cc.
^Mieat starch 8
Wheat flour 169
Wheat bran 342
Eye flour 153
Eye bran . 330
Maize flour 389
Fine-st wheat fl'jur 128
Coarsest wheat flour 486
Mam-izio o holds that the aletu'ometer test is not a trustworthy indication of the value
of a flour for baking, although in general the volume of the bread increases as the
volume of the gluten. He describes the chemical processes of Eobine and Girai-d as
useless, as neither the amount of extractive matter recovered from gluten or flour, nor
the specific gravity of sohiti'jns in acetic acid and alcoholic potash, indicates the value
of a sample to the baker. Xor can fermentation tests replace baking tests. The
specific gravity of the bread is of value, and varies as follows:
Best qtiality 0. 2:3 to 0. 28
Medium quality 0.35
Inferior quality 0. 46
LiebeiTuann c has designed a special apparatus for determining the expansion of
gluten on heating in an oil bath to 170° C. for fifteen minutes. Four samples of flour
using 20 grams; gave as follows:
Erpan-sion of jiour determined hy Liehermann' s apparatus.
No. of Moist Expan- No. of Moist i Expan-
sample. gluten. sion. sample, i gluten. |i sion.
Per cent. cc. i' Per cent.
31.6 135 3 i 24.6
35.2 120 4 i 22.3
Snyder ci asserts that flour of good quality should contain 12 per cent of total proteids,
or 11 per cent protein <X X 5.7 u of which from 55 to 65 per cent should be gliadin.
Xorton, f commenting nT\ the ordinary mf-thods of ascertaining the bread A'alue of
floiu's. says:
Some workers determine only the dry. crude gluten, others only the moist or wet
gluten, whilst others determine both, and express the relation of the two as water
capacity. If a gltiten has a good water capacity with proper physical qualities it
wotild be desirable, but there does not seem to be amthing very definite about the
values obtained, and often a high water content goes with excess of gluten in and poor
bread-making properties, so that the determination does not seem to be of much value.
The admission that a flom may contain excess of glutenin mtist he taken as corrobo-
rative of the conclusions already quoted) reached by Guess and Snyder to the efi'ect
that excess of gliadin in flour is imknown. Yet in a later sentence Norton seems to
aZts. Xahi-. Genussm.. 1905. 10: 747— Abst. Analyst. 1906, SI: 73.
&Landw. Jahrb., 1902, 31: 179— Abst. Analyst. 1902, :27: 249.
cZts. Xahr. Genussm., 1901. -..• 1009— Abst. Analyst. 1902, 27: 155.
dj. Amer. Chem. Soc, 1904. -26: 263— Abst. Analyst. 1904, 29: 157.
ej. Amer. Chem. Soc. 1906. 28: 19.
71
admit the possil)ility of an excess of gliadin, when he says: ' 'On the other hand, if the
separated gluten lacks body, an excess of gliadin is indicated." Norton advises
reporting the gluten as dry gluten, since large personal and other errors enter into the
determination of moist gluten.
On the whole, it would seem to be the opinion of Norton, Snyder, Arpin, and other
investigators that the crude gluten number has little value in fixing the quality of
flour; at least where separation of gluten in the ordinary way is practiced. Fleurent
and Manget use a dilute solution of sodium chlorid for washing out the starch, etc.
The method is troublesome, and Chamberlain « has found it to give too high results.
Macfarlane ^ has recorded results of work on gluten estimation by the following
processes :
The dough balls from 10 grams flour stood for 30 minutes and were then washed with
250 cc distilled water; residual gluten, dried and weighed in one case, but duplicate
gluten balls, were washed with 250 cc of 70 per cent alcohol, dried and weighed. The
resultant weights gave "crude gluten" and "crude glutenin," respectively, and the
difference gave " crude gliadin. "Nitrogen was determined in the aqueous washings,
and calculated to water soluble proteids; and in the alcoholic washings, and calculated
to pure gliadin. The difference between crude gliadin and pure gliadin is designated
by Mr. Macfarlane as " dextrinoids. " The nitrogen of crude glutenin was determined
by the Kjeldahl method, and multiplied by 5.7 to give " pure glutenin." The differ-
ence between this number and the crude glutenin number is styled "nonproteids in
crude glutenin. " The total nitrogen of the flour multiplied by 5.7 gives total proteids.
The proteids found as above described are generally less than the total proteids by
about 1.5 per cent; but in certain flours they show an excess of about 0.5 per cent.
If we denominate as " true gluten " the sum of " pure glutenin " and " pure gliadin, "
as found by Macfarlane, then the "true gluten" found constitutes from about 72 per
cent to 94 per cent of the "total proteids" (NX5.7) of the flour, or from about 70 per
cent to about 90 per cent of the crude dry gluten. The number is evidently not com-
parable with that to which Norton and Chamberlain have given the name "true glu-
ten," and which they found to average 75 per cent of the weight of the crude gluten.
Mr. Macfarlane acknowledges that the intelligible appreciation of his analytical results
awaits further study.
In a second paper read before the Royal Society this year (May, 1906) Mr. Macfar-
lane supplements his work by further analytical results on the same lines. Comment-
ing on the difference found between total proteids as calculated from total nitrogen,
and from the sum of those obtained in analysis, the author says: " This quantity varies
from 0 to 2.52 per cent on the original flour, and may possibly yet afford useful indica-
tions as regards the physical character of the gluten from which it is separated." The
difference referred to generally shows a loss in washing out the starch, which loss is
probably due to glutenin.
Since the wheat crops of 1903, 1904, and 1905 possessed well-recognized characters,
as determined by world-wide baking tests, it has been possible for Mr. Macfarlane to
interpret his results with reference to the l;nown character of the flours. He finds the
most distinctive characteristic to be the gliadin-glutenin ratio, which is 41.07: 58.93 .
for the best quality of flour (crop of 1903), and concludes: "The most advantageous
proportion of gliadin to glutenin for baking purposes is about 40 to 60, it being under-
stood that the proteids removed with the starch in the production of the gluten are to
be regarded as glutenin."
This is entirely at variance with hitherto accepted standards; and, as based upon
t3xtensive and careful work, practically reopens the whole question.
«U. S. Dept. Agr., Bureau of Chemistry, Bui. 81, p. 118.
& Trans. Roy. Soc. of Canada, 1905, [2], 11, Sec. Ill: 17.
72
Gliadin. *
As already mentioned, Snyder « suggests that gliadin as usually separated may not
be a simple proteid.
Konig and Rintelen b have isolated three distinct proteids from the extractive
matter obtained by treating gluten with 65 per cent alcohol. According to these
investigators, gluten would seem to have the following composition:
Gluten :
A. Insoluble in 65 per cent alcohol =gluten casein (glutenin?).
B. Soluble in 65 per cent alcohol —
I. Gluten fibrin — soluble in 90 per cent alcohol,
II. Mucedin — soluble in 40 per cent alcohol.
III. Gliadin — insoluble in 40 per cent alcohol.
The method by which these separations were effected is much too complex to be
available for routine work, and lends probability to a suspicion that changes in the
character of the proximate components of gluten may occur during the manipulations.
Snyder c describes a method for the direct separatioij of gliadin from flour by
shaking with 7 J per cent alcohol, and determination of the gliadin in solution, by
means of the polarimeter. He finds that with 15.97 grams of flour and 100 cc of solvent
the reading in a 220 mm tube at 20° C. gives a number which multiplied by 0.2
yields a product in close agreement with gliadin nitrogen as obtained by the ordinary
Kjeldahl process.
Mathewson^ has recently made a study of the specific rotation of gliadin. He
finds that this is practically independent of the concentration in 70 to 75 per cent
alcohol. With 70 to 80 per cent alcohol it decreases with increase in the alcohol
strength. Increase in temperature within the limits 20°-45° C. produces a slight
increase in the specific rotation. He finds also that the differences in density for
gliadin solutions such as would be met with in ordinary analysis are too insignificant
to allow a margin for experimental error in this method of determining gliadin. The
method was originally suggested by Fleurent. e
Detection op Foreign Starches, Especially Maiz:^ Starch, in Wheat Flour.
Kraemer/ gives special instructions for sampling, mounting, and observing starches,
and records the results of treatment with sixteen different reagents in microchemical
research. While many starch granules in maize and wheat closely resemble each
other, the characteristic forms enable approximate quantitative estimations to be
made. He advises the use of polarized light in this kind of work.
He points out that the odor of a flour is of value in detecting the presence of maize.
Embrey g has found from 10 to 30 per cent of maize in wheat flour, and in " self-
raising" flour from 10 to 20 per cent.
He refers to work by White h on maize in oatmeal, and to Baumann i, who points
out that 1.8 per cent potash ruptures wheat and rye starch granules without affecting
the maize starch, and its action can be immediately stopped by adding acid. This
reagent is otherwise superior to chloral hydrate. Embrey makes the microscopic
a J. Amer. Chem. Soc, 1905, 27: 1068.
& Zts. Nahr. Genussm., 1904,^; 401.— Abst. Analyst, 1904, £9: 371.
c J. Amer. Chem. Soc, 1904, 26: 263.— Abst. Analyst, 1904, 29: 157.
d Ibid., 1906, 28: 624.
e Compt. rend., 132: 1421.— Abst. J. Soc. Chem. Ind., 1901, 20: 941.
/ J. Amer. Chem. Soc, 1899, 21: 650.
^Analyst, 1900, 25: 315.
^ Ibid., 1895, £0.- 30.
i Ibid., 1899, 24: 150.
73
test more exact by the use of iodiii, or by determining sugar after hydrolysis with
sulphuric acid.
In discussion it was brought out that clove oil renders the maize hilum black, while
that of wheat remains invisible. Embrey concludes that only roughly quantitative
results can be obtained by microscopic work.
Volpino « depends upon either an estimation of total gluten, which is much reduced
in quantity by admixture of other flours with wheat; or upon estimation of nongluten
proteids which are mechanically washed out of the flour. These are recovered from
the wash water by asbestos filtration, and nitrogen determined. These proteids
(NX 6) were found as follows:
Percrnt.
Wheat (lour less than 0. 2
Maize flour less than 6. 5
Barley and rice fiourj CO
Rye flour 5. 0
Leroy^ finds that phoroglucinol colors sawdust much more distinctly than it does
the cellulose of flours. Quantitative interpretation of results under the microscope is,
however, very unsatisfactory.
Balland c has found that the fatty matters of freshly milled flour consist of a very
fluid oil and solid fatty acids of different melting points. In the course of time the
acidity diminishes and fatty acids increase, the ratio of increase (up to a certain point)
being a measure cf the age of the flour.
The fatty acids themselves disappear in time and are not found in very old flour.
The acidity which is the first indication of alteration of the flour, is not connected with
bacterial decomposition of the gluten, but is derived directly from the fat, and occurs
in the fat when this is extracted by ether. The gluten is not attacked until the fatty
acids produced from the oil begin to disappear. The richer the flour is in oil the more
liable to alteration, e. g., flour from hard wheat.
Gudeman^ points out that the fat in maize flours can not be accurately estimated
from a dry gluten mass, as the fat is altered in drying. He prefers rosolic acid as an
indicator in determining the acidity of flours.
Watkins e shows that a distinct acidity in bread is needed to prevent the develop-
ment of Bacillusifniesentericus, which gives rise to the disease called "ropiness. " The
natural acidity of bread not being sufficient to hinder this development, he suggests
the addition of acetic acid in baking, under certain circumstances.
EEPORT OF COMMITTEE 0 ON EEOOMMEI^DATIONS OP EEPEKEES.
By H. C. Lythgoe, Chairman.
(1) Colors.
It is recommended —
1. That, as suggested by the associate referee for 1905, the cooperative work on colors
be continued along the following lines:
(a) Solubility of the coal-tar and vegetable dyes in various solvents (ether,
acetic ether, petroleum ether, methyl and ethyl alcohols, acetone, etc., and mixtures
thereof), arranged according to their solubility — as, easily soluble, difficultly soluble,
and insoluble.
aZts. Nahr Genussm., 1903, 1089— Abst. Analyst, 1904, 29: 89.
&Chem. Ztg., 1898, 22: 311; the Analyst, 1900, 25: 39.
cCompt. rend., 1903, 724.— Abst. J. Soc. Chem. Ind., 1903, 22: 1303.
dV. S. Dept. Agr., Bureau of Chemistry, Bui. 73, p. 42.— Abst. J. Soc. Chem. Ind.,
1903, g^.- 764.
cj. Soc. Chem. Ind., 1906, 25: .350.
74
(h) Extractive values of the various solvents for dyes in neutral, acid, and alkaline
solutions.
(c) Cliai-acteristics of the coloring matters as contained in fi'esh fruits, vegetables,
wines, etc.. with reagents and solvents, including their respective dyeing properties.
Adopted.
2. It is the sense of the committee that the term "• vegetable colors" is a misnomer
as applied to the products studied by 'Mr. Xickles. and the committee recommends
that this portion of the report be retm-ned to the referee with the stiggestion that the
words "commercial colors alleged to be vegetable colors"' be substituted for the words
•''vegetable colors"' and that this portion of the report be revised accordingly.
Adopted.
(2) Saccharixe Products.
The referee made no formal recommendations, but suggested that if the work on
maple products was continued the thi-ee methods tried for the determination of the
lead ntimber be given a thorough trial [see page 18]. and,that the work on the malic
acid value be continued imtil a modification is secured which gi^-es more uniform
results in the hands of different chemists.
Xo action was taken by the committee on these suggestions, which are submitted
for the information of the referee for 1907.
> 3 1 Distilled Liquors.
It is recommended that a further study be made of the determination of fusel oil.
Xo action taken by committee, and recommendation goes to the referee for consid-
eration.
(4"i Beer.
The following recommendation was not acted upon by the committee, as it involves
action as to an official method, and is referred to the referee for 1907 for recommendation
as to final action, together with the suggestion made as to the reA'ision of thealcoholo-
metric tables:
I suggest that a committee be appointed by this association for the pm'pose of revising
the alcoholometric tables now given in Bulletin Xo. 65, that they may be adopted for
use at this new temperattue (20° C. ). I furthermore recommend the adoption of the
following methods»of beer analysis as official for this association: [See Circular Xo. 33,
Btueau of Chemistry, for detailed statement of methods.]
(5) IXFAXTS" AXD IXVALIDS" FoODS.
It is recommended that the subject "Infants" and Invalids" Foods" be dropped.
Mr. Bigelow explained tiiat it was thoiiglit best to drop this division
of the work for the reason that it was a diipHcation, nearly all infants'
and invalids' foods being included under cereal products, meat
extracts, and dairy products.
Mr. TTiley indorsed the view that in general these foods are only
general foods which should not be permitted to bear the character-
istic name of infants' or invalids' foods, and also called attention to
the effort making in the Council of Pharmacy and Chemistry of the
American Medical Association to classify medicinal foods; that is,
foods never used except in cases of illness or convalescence.
The association accordingly voted to adopt the motion discontin-
uing the subject as a separate investigation.
75
(6) Condiments othkk tiiax Spices.
It is recommended that the methods outlined by the referee [see page 39] be adopted
provisionally, but that in printing the same they be given by reference wherever such
methods appear under other subjects.
Action deferred for one year.
(7) Tea and Coffee.
It is recommended —
1. That the methods outlined by the referee for the estimation of caffetannic acid
[see page 44] be submitted for cooperative study by the members of the association.
Adopted.
2. That the Gomberg method for the determination of caffein be subjected to trial.
[See page 45.]
Adopted.
(8) Food Preservatives.
Action on recommendations 1-7 [see page 57] involving changes in the provisional
methods in Bulletin No. 65 was deferred until 1907.
The four recommendations as to work for the ensuing year [see page 58] were adopted.
(9) Determination of Water in Foods.
It is recommended that the study of the vacuum method for the determination
of water be continued next year and that the methods employed for study shall be:
(a) The present vacuum method in either the Hemple or Scheibler type of desiccator;
(b) the vacuum method with any modification which may facilitate drying, as the
introduction of phosphorus pentoxid as a drying agent; (c) drying in partial vacuum
in a current of dry preheated air and inert gases.
Adopted.
(10) General Recommendations.
The committee recommends that the associate referees on the following subjects
study the determination of alkalinity of soluble and insoluble ash with a view to
obtaining uniform results: Saccharine products, including confectionery; fruit pro-
ducts; wine; beer; vinegar; spices; dairy products (in reference to calcium sucrate
in cream); cereal products; vegetables; cocoa and cocoa products; tea and coffee.
Adopted.
At the close of the report of Committee C, Prof. C. E. Munroe, of
the George Washington University, welcomed the association to the
university halls and in a brief speech invited their cooperation in
insuring the success of the proposed exhibit at the Jamestown
Exposition, 1907, of the applications of denatured alcohol. Pro-
fessor IMunroe, who has charge of the exhibit, said that its object
was educational, and made the following statement in regard to the
same :
By act of Congress approved June 7, 1906, alcohol which has been denatured, or
rendered unsuitable for drinking purposes, may, from and after January 1, 1907, be
used in the arts and industries, and for fuel, light, and power, without the payment
of internal-revenue tax, thus making a cheap and inexhaustible source of energy
available to our people with which to supplement the coal, petroleum, and gas deposits
76
which are now being too rapidly drawn upon. To enable our people the more readily
to avail themselves of the legislation and to promote invention, through which the
largest advantage may be taken of this privilege, an exhibit will be made of the vari-
ous apparatuses, machines, and appliances by which the heat, light, and power
which can be obtained from alcohol may be utilized for domestic, agricultural, and
manufacturing purposes, and also of the articles of manufacture into which alcohol
entere as a component or factor.
In tliis connection Doctor Wiley laid before the convention sev-
eral invitations received from the authorities of the exposition invit-
ing the association to hold the meeting of 1907 at Xorfolk under
their auspices. Action as to the acceptance of these inWtations
was deferred.
The report on nitrogen was then presented by the secretary on
behalf of the referee.
EEPOET 0¥ I^ITROGEIT.
By James H. Gibboxey. Referee.
The work of the referee for the past year is embodied in the following recommenda-
tions of the association adopted at its last annual meeting:
(1 ) That the study of availability by both the netural and the alkaline permanga-
nate methods be continued, particularly with the A"iew to determining the amount of
material to be used in the netitral method on commercial fertilizers.
(2) That in the official Gtmning method for the determination of nitrogen the addi-
tion of 0.5 to 0.75 gi-am of copper sulphate after digesting with sulphtnic acid and
potassium sulphate for one half horn- be studied by the referee. (Fuller.)
(3) That the silver precipitation method (Bui. '46, Rev., p. 14), under "'4. Deter-
mination of nitrogen," be changed to read as follows: (See Penny's modification,
page 77.)
The two samples selected for the nitrogen work were prepared from high-grade acid
phosphate, dried blood, and cotton-seed meal: Xo. 1, acid phosphate and dried blood;
Xo. 2, acid phosphate and cotton-seed meal.
The materials were finely gi'ound and au* dried at temperattne of laboratory tmtil
determinations proved moisture content to be constant. They were then mixed to
give approximately a 2 per cent nitrogen content. The average ci six determina-
tions of Xo. 1 (maximtim of 2.10, a minimum of 2.08) was 2.09 per cent nitrogen, while
that of X.o. 2 gave (maximum 2.16, minimum 2.12) average of 2.14 per cent nitrogen.
The solution for hydrochloric acid determination, approximately half normal, was
prepared by using freshly distilled water and strictly chemically pure acid.
In response to a circular letter asking for cooperation. 36 f^^vorable replies were
received from as many laboratories. Sixteen of thes: sent results which represented
the work of 26 chemists.
The following instrtictions accom-nanied each s-:^t of samples:
IXSTRUCTIOXS FOR XITROGEX WORK. 1906.
The work on nitrogen for the Association of Official Agricuhural (Chemists, in which
yoti have expro^^-sed your desire to cooperate, include? the following determinations:
TrAaJ nitrogen.
(1 ) Determine total nitrogen by Kjeldahl or Gtmning methods.
(2) Determine total nitrogen by Gunning method.
('3 I Determine total nitrogen by the following modification of the ofiicial Gunning
method ;
77
Ftdlcr^s modification. — In a digestion flask holding 500 cc place from 0.7 to 3.5 grams
of the substance to be analyzed, according to its proportion of nitrogen. Then add
10 grams of powdered potassium sulphate and from 15 to 25 cc of concentrated sulphuric
acid. Conduct the digestion as in the Kjeldahl process, starting with a temperature
below the boiling point and increasing the heat gradually until frothing ceases. After
the digestion has continued for 30 minutes remove the flame and add cautiously
0.75 gram of powdered copper sulphate and continue the digestion to the end. Dilute,
neutralize, and distil as in the Kjeldahl method.
Make one determination following above method and one in which 10 cc of strong
potassium sulphid s(^lution is added just after diluting, neutralizing, and distilling, as
in Kjeldahl method.
Availahle nitrogen.
(1) Neutral permanganate method. — (a) Using a charge corresponding to 0.075 gram
nitrogen. (6) Using a charge of 2 grams sample.
Into a 300 cc low form Griffin beaker weigh 2 grams of the sample if from a mixed
fertilizer; if from a concentrated material use a quantity containing approximately
0.075 gram nitrogen. Samples containing materials that have been treated with acid
should be washed on a 9 cm S. S. No. 595 filter to 200 cc and transferred, filter and all,
to beaker. Digest this with 125 cc of permanganate solution (16 grams of pure potas-
sium permanganate to 1,000 cc water) in a steam or hot-water bath for 30 minutes.
Have the beaker let down well into the steam or hot-water and keep closed with a
cover glass, stirring twice at intervals of 10 minutes with a glass stirring rod. A* the
expiration of the time remove from bath, add 100 cc of cold water, and filter through a
heavy 15 cm folded filter. Wash with cold water, small quantities at a time, until
total filtrate amounts to 400 cc. Dry and determine nitrogen in residue by Kjeldahl
method.
(2) Alkaline permanganate method. — Weigh out an amount of sample containing 0.045
gram of nitrogen, and transfer to a 600 cc distilling flask. After connecting with
condenser to which the receiver containing the standard acid has been attached
digest with 100 cc of alkaline permanganate solution (16 grams of pure potassium
permanganate and 150 grams of sodium hydrate dissolved in water and made to 1,000
cc) for 30 minutes below the boiling point. Then boil until 85 cc of distillate is
obtained. If the material shows a tendency to adhere to the sides of the flask an
occasional gentle rotation is necessary during distillation.
Determination of hydrochloric acid — Standard solution, official method.
See Bulletin 46, Kev., p. 14, under "4. Determination of nitrogen." Make dupli-
cate determinations by this method.
Penny's modification.— T^j means of a preliminary test with silver nitrate solution,
to be measured from a burette, with excess of calcium carbonate to neutralize free
acid and potassium chromate as indicator, determine exactly the amount of nitrate
required to precipitate all the hydrochloric acid. To a measured and also a weighed
portion of the standard acid add from a burette one drop more of silver nitrate solu-
tion than is required to precipitate the hydrochloric acid. Heat to boiling, cover
from the light, and allow to stand until the precipitate is granular. Then wash with
hot water through a Gooch crucible, testing the filtrate to prove excess of silver
nitrate. Dry the silver chlorid at 140° to 150° C.
Make duplicate determinations by this modification.
Report results under "Total nitrogen" and "Available nitrogen" as per cent
nitrogen.
Report results under "Determination of hydrochloric acid" as gram hydrochloric
acid per cubic centimeter.
We would be glad to have you comment freely on the working of the different
methods, which comments will be incorporated in our report.
James H. Gibboney, Referee.
C. L. Penny, Associate Referee.
The results of tlie deterniinatioiis made bv the different analysts
appear m the following tables:
Table 1. — Total nitroq en.
Analyst.
Official Kjel-
dahl method.
Official Gun-
ning method.
Modified Gun-
ning method
(CuSO,).
Modified Gun-
ninff method
(CuS0^4-
K2S.
1
2
1
2
1
2
1
2
Lewis L. LaShell, Wooster, Ohio
S. H. SMeb. Richmond. Va
W. D. Cooke. Eichmond. Va
Per ct.
i 2.06
1 2.09
2.07
2.05
2.08
1 2.0S
t 2.07
r 2.10
\ 2.12
r 2.09
j:::::::
f
Per ct.
2.18
2.18
2.08
2.09
2.10
9.12
iilO
2.18
2.18
2.14
2.19
Perct.
2.03
2.06
2.14
2.12
2.12
2.08
2.10
2.09
2.10
2.07
2.10
2.19
2.23
2.13
2.15
2.25
"l.'os"
2.09
2.09
2.10
2.11
2.12
Per ct.
2.11
2.13
2.15
2.17
2.16
2.14 1
2.14
2.20
2.24
2.16*
Per ct.
2.10
2.11
2.10
2.14
2.12
2.12
Perct.
2.14
2.12
2-12
1 2.15
-2.15
2.n
Per ct.
2.16
2-16
2.11
Per ct.
2.16
2.18
2.15
F. B. Carpenter. Eichmond. Va
W. P. .men. New Brunswick. X. .J..
"'2."i2'
2.'i7
John Phillips Street. Xew Bruns-
wlck. X. J
2.07
2.04
2.07
2.27
2.23
2.17
J. 0. Mundy, Blackshurg. Va
•J. H. Xorton. Favettevine. Ark
2.2i 1
2.26
2.25
2.13
2.18
2.05
""2."i2'
2.21
•? 19
T. C. Trescot. Washington. D. C [
2.15 i
2.13 t
2.15
2.23
2.25
2.16
2.16
2.18
2.19
2.10 1
2.13
2.13
2.12
A. W. Blair. Lake Citv. Fla
f 2.00
\ 2.05
1 2.13
f 2.09
1 2.08
1 2.08
I 2.10
r 2.14
{ 2.15
1 '' 12
1 2!09
2.05
2.08'
2.18
2.14
2.16
2.14
2.15
2.11
2.10
2.09
2.13
2.11
James H. Gibhoney, Roanoke. Va...
Geo. A. Olsen and F. W. WoU.
Madison, Wis
2.10
2.11
2.13
2.10
2.11
2.15
. 2. is
2-17
2.18
i 2.17
2.1s
2.07
2.08
2-10
2-09
2-10
2.15
2.10
2.16
2.15
2.17
2.17
2.10
2.0s
L. Eosensteiu. Berkeley. Cal •
2.09
2.10
2.20
2.16
■' 2.11
2.12
2.24
2.24
2.04
2,11
2.05
2.15
f 1.99
I L99
2.16
2.18
2-18
2.16
2.23
2-24
2.33
2.13
2.08
, 2.11
i 2.17
2.22
1 2.30
I 2.13
2.13
2.15
1 2.13
'iio'
2.09
2.09
2.08
""2." if
2.14
"i'os'
2. 18
2.16
2.1s
2.1s
E. B. Holland and P. H. Smith.
Amherst. Mass
Samuel W. Wiley and W. E. Ho3-
man, Baltimore. Md
W. D. Eichardson. Chicago. IlL
B. F. Robertson, Clemson. S. C
/ 2.06
i 2.06
f 2.08
i 2.10
r 2.13
t 2.10
f 2.07
I 2.07
2.i6
2.15
2.14
2.16
2.17
2.25
2.13
2.16
2.14
2.15
2. is
2.08
2.09
2.11
2.13
2.35
209
2.12
2.10
2.10
2.16
2.08
2.02
2.00
2.06
2.06
"2. 08'
2.06
2.36
9.22
2' 13
2.15
2.1s
2.47
2.13
2.08
■ 2.14
2.13
2.20
2.20
2.0s
2.06
i 2.03
2.05
2.14
2.17
"iii"
2.09
2.04
2.02
2.13
2.19
■->->2
■im
2.11
"2.64'
2.09
2.18
2.08
1.95
2.01
2.02
""2.' 06'
2.08
2.08
2.10
2.13
2.19
2.21
2.09
2.12
2.13
2.17
2.21
2.2s
C. C. McDonnell, Clemson. S. C
2.03
2.02
"iio'
2.08
2.06
2.12
D. H. Henry, Clemson. S. C
Jerome J. Morgan. College Park.
Md s
2.08
j:::::::
2.12
........
2. is
2.10
2.06
""2." 64'
C. H. Jones. Burlington. Vt
R. W. Thatcher. Pullman. Wash
f 2.07
i 2.07
[ 2.07
/ 2.08
\ 2.09
2.11
2.14
2.14
2.14
2.15
2.14
2.17
■""2." OS
Table 2. — Availabh nitrogen, neutral ptrraanganate method.
[Two grams 01 sample=0.075 gram nitrogen.]
Analvst.
Insoluble
nitrogen.
Available
nitrogen.
Insoluble
nitrogen.
Available
nitrogen.
Per ct. Per ct. Per cU Per cL Perct. Perct. Perct. Perct.
1.25
W. D. Richardson. Chicago. lU...
L. Rosenstein, Berkeley. Cal | • ** j
E. W. Holland, Anxherst. Mass.
Samuel W. Wiley and W. E. Hofi- [
man. Baltimore. Md 1
0.36 41.59 S4.O0
.51
.55 |! 81.06 74.54
.20 58.24 90.61
.16
.17
.20 So. 17 95. S3
1.32
0.45
as. 31
.26
SO. 00
79
Table 2. — Available nitrogen, neutral jxrmanganate method — Continued.
Analyst.
Insoluble
nitrogen.
Available
nitrogen.
Insoluble
nitrogen.
Available
nitrogen.
1.
2.
1.
2.
1.
2.
1.
2.
Per ct.
Per ct.
Per ct.
Per ct.
Per ct.
Per ct.
.29a
.43b
.39c
.18d
.16e
.259
.239
.314
.308
.12
.13
Per ct.
\
87.74
Per ct.
.18
.21
.19
.18
f .34
\ .32
f .285
i .41
.16
.14
.16
.18
.19
.17
.18
.17
.24
.26
82.04
Jerome J. Morgan,College Park, Md . .
91.75
91.18
92.23
""'.'459"
.417
.943
.870
.43
.43
W. P. Allen, New Brunswick, N. J. . .
Jolin P Street New Brunswick,
84.30
91.24
78.94
88.09
N.J
83.53
91.97
57.01
85. 74
S. H. Shieb, Richmond, Va
1
79.91
93.95
i .93
i .95
.03
r .32
\ .31
f .86
t .64
.31
.34
' .36
/ .34
i .26
.17
.17
.20
.29
.24
Lewis L. La Shell, Wooster, Ohio. . . .
54.59
70.56
92.20
90.65
T. C. Trescot, Washington, D. C
J. E. Greaves, Logan, Utah
84.43
88.01
A. W. Blair, Lake City, Fla
"'""."is'
.17
.18
.23
.23
63.. 59
.39
.41
.44
.46
.46
.23
'.24
.27
83.73
92.57
80.3^
88.38
R. W. Thatcher, Pullman, Wash....
85.58
89.25
77.84
Table 3.^ — Available nitrogen, alkaline permanganate method.
[Amount taken=0.045 gram N.]
Analyst.
Volatile
nitrogen.
Available
nitrogen.
W. D. Richardson, Chicago, 111
L. Rosenstein, Berkeley, Cal
E. W. Holland, Amherst, Mass...
P. H. Smith, Amherst, Mass
James H. Gibboney, Roanoke, Va.
R. W. Thatcher, Pullman, Wash. .
Samuel W. Wiley and W. E. Hoffman, Baltimore, Md.
Jerome J. Morgan, College Park, Md
C. H. Jones, Burlington, Vt
S. n. Shieb, Richmond, Vi
Lewis L. La Shell, Wooster, Ohio.
T. C. Trescot, Washington, D. C.
J. E. Greaves, Logan, Utah
J. H. Norton, Fayetteville, Ark.
A. M\ Blair, Lake City, Fla
cent.
0.87
1.17
1.12
1.27
1.24
1.28
1.31
.84
.88
1.02
1.06
1.10
1.32
1.29
1.31
1.29
1.29
1.31
1.27
1.25
1.26
1.11
1.14
1.25
1.23
1.14
1.16
1.16
1.45
1.48
1.32
1.31
Per cent.
1.17
.87
.80
.95
.92
1.02
1.04
.84
.84
.69
.72
.75
1.05
1.00
1.03
1.03
1.03
1.00
.98
.96
.99
1.36
1.48
1.12
Per cent.
40.65
Per cent.
52.00
57.28
60.18
58.76
39.63
44.60
43.19
62.20
47.91
"39 .'25
33.33
50.00
"47 ."69
58.88
45.35
1.41
1.04
1.15
1.13
.54.59
58.41
65.14
52.34
58.79
66.51
42.38
63.80
63.59
51.43
80
Table 4. — Deterinination of hijdrochloric acid.
[Gram per cubic centimeter.]
Analj'st.
Offci 1
method.
( 0.0153659
.015^396
John Phillips Street. New Brunsviick, N.J . 0153710
.015.142
I .0153 38
J. H. Norton, Fayetteville, Ark ;{ • ^Jp^
T.- C. Trescot, Washington, D. C I a. 015229
G. A. Olsen and F. W. Woll, Madison, Wis [ • §}g
Samuel W. Wilev and W. E. Hoffman, Baltimore, Md '. 0152535
W. D. Richardson, Chicago, 111 .01524
f .01535
James H. Gibboney, Roanoke, Va | • ^J^^j
1 1 ; 01537
Jerome J. Morgan, College Park, Md , { -^jgl
L. Rosenstein, Berkeley, Cal j • q\^1^
Modified
method.
0. 01530
"6."6i5i225
.0153
. 0154
. 0152405
. 01507
.01533
. 01537
. 01539
.01536
. 01532
.01532
. 01522
^Average of 3 determinations. f> Average of 2 determinations.
COMMEXTS BY ANALYSTS.
/. H. Norton: In the alkaline permanganate method digestion continued only 15
minutes as the solution unexpectedly boiled and it was feared that the standard acid
would strike back if the temperature were lowered.
/. E. Greaves: In the alkaline permanganate method on sample 2 the results ran
0.92 to 1.226 per cent, and agreement between the duplicates could only be
obtained by distilling off exactly 85 cc. Even after this ammonia is given off until the
substance is quite dry.
Jerome J. Morgan: In digesting the sample (2) in determinations a. b. and c with
125 cc neutral permanganate solution as directed it was found that the solution was
decolorized. This led to the trial of using 200 cc of the permanganate sohition. vrhich
gave the results in determinations d and e. These results check closely with the results
on the same sample, using a 2-grain charge.
John Phillips Street: Mr. W. P. Allen and myself both found that bumping
caused much trouble when the modified Gunning method was used. Our availabe
results were quite satisfactory except on sample 1 when 0.075 gram nitrogen was used,
the two sets of results not agreeing with each other at all. We followed your directions
in all particulars and in the availability tests washed with 200 cc cold water before
adding the permanganate. In standardizing the hydrochloric acid sent, the acid was
measured in five 10-cc portions from a 50-cc burette. This burette was marked as
having been calibrated, but the results secured caused somie doubt as to the accuracy
of the calibration. You will notice that the results vary from 0.6043 to 0.6072 gram
silver chlorid, too wide a variation to be satisfactory.
S. H. Shieb: The copper sulphate addition in the modified Gunning method is
objectionable in that it deposits copper on the zinc during distillation, interfering
with the evolution of hydrogen, and thereby causing bumping. The addition of potas-
sium sulphid does not appear to prevent this.
L. Rosenstein: In the hydrochloric acid standardization a 5 per cent solution of silver
nitrate was used in precipitating the chiorin.
C. C. McDonnell: A conclusive opinion can not be reached from work on two
samples, but from this limited number of determinations the methods in which copper
sulphate, and copper sulphate and potassium sulphid were used show no advantage
over the official method. In fact the results are more variable and the method slightly
more troublesome.
81
Geo. A. Olsen and F. W. Woll: Temperature of hydrochloric-acid solution when
drawn from burette, 2Q° C. All silver chlorid precipitates were dried at 130° C, the
highest temperature that could be reached in the bath used.
T. C. Trescot: As you will observe, the results by the modified method (determina-
tion of hydrochloric acid) are decidedly lower than those by the official method. In
washing with hot water in the modified method I was unable to reach a point where
the silver chlorid ceased to dissolve in the wash water.
E. B. Holland and P. H. Smith: The addition of copper sulphate in the modified
Gunning method, after one-half hour digestion, proved rather a troublesome feature
for commercial testing, and in distilling frothed badly, especially where no potassium
sulphid was added. Digested two hours.
Neutral permanganate method: This process requires too much time and attention
for one in which the results are merely indicative at best. The retarding action
of the filter paper during digestion and washing, together with variations in digestion
emperature (with our facilities), tends to prevent concordant results.
Alkaline permanganate method: This method is short and simple, but absolutely
uniform conditions of digestion and distillation are necessary to obtain agreeing
results. Contrary to directions, the time of digestion at this station is taken from the
beginning of the ebullition. Distillation almost to dryness (15 cc) is objectionable
and often introduces an error. If the quantity of solution were increased to 150 cc (or
200 cc) and 100 cc (or 150 cc) distilled, digestion taken from time of ebullition and
perhaps increased to one hour, it would tend toward higher and probably more uniform
results.
Comments by Referee.
The modification of the Gunning method suggested by Mr. Fuller was found to be
of no advantage, the yield of nitrogen being within a few tenths per cent of the official
Gunning method. The difficulties presented by the addition of copper sulphate were
two — loss of time in cooling solution down to add the salt and bumping of solution dur-
ing distillation. This last point was very objectionable and was not decreased to any
great extent by the addition of potassium sulphid to remove the copper from solution.
The neutral permanganate method was found to be very tedious and difficult to con-
trol. It is full of seemingly insignificant details which when varied cause great varia-
tions in the results. For instance, changes in results were obtained by varying the
length of time of stirring at the two intervals, by variation in temperature of water bath
during digestion (a condition which very often happens unnoticed), length of time
required for filtering the digested residue from the permanganate solution, method of
washing, and time consumed in washing. The results obtained were too high as com-
pared with availability by actual experiments. From experience with this method
relative to quantity of material to be used in commercial fertilizer work I am of the
opinion that the proportion of nonnitrogenous organic matter to nitrogenous organic
matter should be the determining factor. Jn the case of sample 1, the charge which
gave results nearer field experiments was the one corresponding to 0.075 gram, while
in that of sample 2 the charge of 2 grams was found to be nearer the field availability.
It would be a very difficult matter indeed to have a sliding variation of quantity of
material to betaken for the many materials used as sources of nitrogen. Lack of time
l)revented a more definite study of this point.
The alkaline permanganate method was found to be much easier to manipulate and
control, the only points requiring strict attention being to keep the digestion just
below the boil for the thirty-minute period and distilling off the exact 85 cc of solu-
tion. The first point was easily controlled after several preliminary tests. The second
point always gave more or less trouble, it being very difficult to distil to such a low
point. It was found an easy matter to get good duplicates when distillation wag
31104— No. 105—07 6
82
stopped at 85 cc, but, as has often been pointed out, continued distillation after the
85 cc portion will give appreciable quantities of ammonia. Several determinations
were made by adding 25 and 50 cc of water just after digestion and distilling 100 cc.
Very little variation was found by this procedure.
The standardization of hydrochloric acid by both the official and the modified
methods was very satisfactory. No variation was found by the two methods, if the
modification can really be called a modification of the official method. This prelimi-
nary test to determine the amount of silver nitrate solution within a drop or two
was found to be of great value, as less washing was required to remove the small excess
of precipitant. Continued washing with hot water was found to be very objectionable,
in that very appreciable quantities of silver chlorid were dissolved, passing into the
washing. Continued heating of solution during precipitation to collect precipitate
was also found to cause a slight variation, due to dissolved silver chlorid. It was
found to be very desirable to make the precipitation in a small Erlenmeyer flask, and
after adding the desired amount of silver nitrate to wrap in towel and shake vigorously
for five minutes. The precipitate was in all cases found to be well collected and the
supernatant liquid much clearer than in regular procedilre. In all cases I was very
careful to keep the volume of my solution small and in no cas3 was it necessary to
use 200 cc of hot wash water to remove excess of precipitated nitrates.
Summary.
(1) The Fuller modification, the addition of copper sulphate, and the modification
with copper sulphate and potassium sulphid give no increase in yield of nitrogen and
introduce several very troublesome features.
(2) The neutral permanganate method gave very variable results in the hands of
the different analysts, but with a majority the results were within a reasonable limit.
It will appear that in case of sample 1, in which the nitrogen source is dried blood
(which contained about 15 per cent nitrogen), the amount of material to be used should
be the charge corresponding to 0.075 gram nitrogen. \^Tien the 2-gram charge was
used the amount of nonnitrogenous organic matter was in much smaller quantity,
leaving the permanganate solution stronger and hence more active than where the
charge was increased to represent 0.075 gram nitrogen. With sample 2 the reverse
was found to be true. The cotton-seed meal contained about 6 per cent nitrogen,
hence when a charge corresponding to 0.075 gram was used a greater part of the strength
of the permanganate was used up by the nonnitrogenous organic material, materially
reducing the activity of the solution and thereby decreasing the availability per-
centage.
The results by the alkaline permanganate method compare very closely and, while
somewhat low, the method presents greater possibilities than the neutral procedure.
The results obtained in the standardization of hydrochloric acid were very satisfac-
tory. In no case was there a variation among different analysts that could not be
explained either by the method of drawing sample from burette or variation in cali-
bration and different temperature of liquid when drawn from burette. The modifica-
tion no doubt served a good purpose in determining almost the exact amount of silver
nitrate solution required.
Recommendations.
It is recommended :
(1) That the work on the Fuller modification of the official Gunning method be
discontinued and that the Gunning method remain as it is now stated.
(2), That the work on the neutral permanganate method be continued along the
same lines as this year, having in mind the influence of excessive amounts of non-
nitrogenous material in the source of nitrogen; that work be directed along the line
83
of eliminating s;)nie of the many details of the method which influence to too great a
degree the results obtained.
(3) That the work on the alkaline permanganate method be continued and that the
(piantity of material taken be changed to 0.0675 gram nitrogen; that the quantity of
alkaline permanganate used in digestion be changed to 150 cc, and that 100 cc be
distilled off before titration.
The modified method should read:
Weigh out an amount of sample containing 0.0675 gram of nitrogen and transfer to
a 600 cc distilling flask. After connecting with condenser to which the receiver con-
taining the standard acid has been attached, digest with 150 cc of alkaline perman-
ganate solution (16 grams of pure potassium permanganate and 150 grams of sodium
hydrate dissolved in water and made to 1,000 cc) for thirty minutes below the boiling
point. Then boil until 100 cc of dis'illate is obtained. If the material shows a ten-
dency to adhere to the sides of the flask, an occasional gentle rotation is necessary dur-
ing distillation.
(4) That the Penny modification be added as a footnote to "4. Determination of
nitrogen," standard acid solution.
That the method for standardization be changed in the following points:
(a) That the amount of sjlution taken for determination should be 10 cc, drawn
from burette at 20° C, for half normal solutions, and proportionate amounts for stronger
or weaker solutions.
(b) That the amount of water used in washing silver chlorid precipitate be changed
from 200 cc to an amount necessary to eliminate excess of reagent.
(c) That the precipitation be made in a small Erlenmeyer flask so that after pre-
cipitation the contents of flask can be shaken vigorously for five minutes and the
solution filtered immediately on settling.
THE DETECTION OF PEAT IN GOMMEEOIAL FEETILIZEKS.
By John Phillips Street.
Recently several manufacturing plants have been established for the drying and
pulverizing of peat, and it is claimed that it is the intention to use it as a drier in
mixed fertilizers, and as a diluent of dried blood. While this use of peat may serve
a valuable end in improving the mechanical condition of fertilizers, it also offers a
temptation to the manufacturer to depend upon it, at least in part, for his nitrogen
supply. Recent analyses of 123 samples of New Jersey peat show a range of from
0.74 to 2.83 per cent of nitrogen, with an average of 1.75 per cent. The inertness of
peat nitrogen is well established, and the possibility of its utilization in the com-
pounding of commercial fertilizers is, therefore, a matter of considerable importance.
Von Feilitzen and Tollens « have shown that pentosans are quite generally present
in peat. Fifteen samjdes analysed by them contained from 2.65 to 12.75 per cent
of pentosans, those taken from the upper and less decomposed layers containing
the highest percentages. These high percentages were as a rule accompanied by
relatively high percentages of nitrogen. These facts suggested to the writer the
possibility of employing the phloroglucin method, as used for the determination of
pentosans in cattle feeds, as a means of detecting the presence of peat in com-
mercial fertilizers.
aBer. d. chem. Ges., 1897, 30: 2571.
84
Seven samples of peat, differing widely in nature and mechanical condition, were
analysed Avith the following results:
Analyses of seven samples of peat.
Sample
No.
Total
nitrogen.
i
Pentosans.
Sample
No.
JrogtJp-'—
1
2
3
4
Per cent. Per cent. ]
0.91 ' 2.32 !
1.21 1.57
2.05 2.91
2.14 \ 8.74
5
6
Per cent. | Per cent.
2.65 1 4.98
2.72 ! 2.57
2.98 1 3.56
i
The pentosans ranged from 1.57 to 8.74 per cent, with an average of 3.81 per Gent,
a high percentage of nitrogen and pentosans in general being associated together.
The only material commonly used as a source of organic nitrogen in fertilizers,
which might supply a considerable quantity of pentosans, is cottonseed meal, and
this material has a very limited use in fertilizers in the northern and eastern mar ets.
Average samples of tankage and dried blood were tested for pentosans, and were
found to contain 0.75 and 0.47 per cent, respectively; this was in all likelihood par-
tially du .0 an accidental contamination in both cases. Besides, the official method
for pento^ns is hardly applicable where such small quantities of furfural are obtained ;
3 grams of the tankage yielded only 0.0204 gram of phloroglucid and the same amount
of blood yielded 0.0106 gram. The pentosan determination apparently gave high
results with these two fertilizers, and it is doubtful if either contains more than a
trace.
Three mixtures were prepared as follows:
Mixture No. 1. One part each of nitrate of soda, dried blood, and tankage; 5 parts
of acid phosphate, and 2 parts of muriate of potash. It contained 3.27 per cent of
nitrogen (1.57 as nitrates); 9.42 per cent of phosphoric acid. 10.09 per cent of potash,
and 0.12 per cent of pentosans.
Mixture No. 2. One part each of nitrate of soda, dried blood, tankage, and peat
(No. 5); 4 parts of acid phosphate, and 2 parts of muriate of potash. It contained
3.53 per cent of nitrogen (1.57 as nitrates'), 7.98 per cent of phosphoric acid, 10.09
per cent of potash, and 0.62 per cent of pentosans.
Mixture No. 3. One part each of nitrate of soda, dried blood, and tankage; 3 parts
of peat; 2 parts of acid phosphate, and 2 parts of muriate of potash. It contained
4.06 per cent of nitrogen (1.57 as nitrates^ 5.10 per cent of phosphoric acid; 10.09
per cent of potash, and 1.62 per cent of pentosans.
These mixtures represented in composition high-grade fertilizers, the first con-
taining no peat, the second 10 per cent of peat, and the third 30 per cent. It is doubt-
ful if as small a quantity of peat as 10 per cent would be used in actual practice, but
it was used here to test the delicacy of the method.
Pentosans were determined in each of the mixtures by the phloroglucin method.
No. 1 yielded a trace; No. 2, 0.52 per cent, and No. 3, 1.41 per cent. These figures
are slightly below theory, but show very fair agreement when the difficulties of the
method are taken into consideration. It is seen, therefore, that a pentosan determi-
nation will indicate as small an addition as 10 per cent of peat. If cottonseed meal,
or castor pomace were used to compound the fertilizer, the test for pentosans would
not establish with any certainty the use ol peat, for these materials both contain
pentosans. However, these materials are used but little in the northern States
(excepting castor pomace in tobacco fertilizers), and it is believed that this simple
test will in the great majority of mixed fertilizers indicate whether or not peat has
been used; it will likewise serve for the detection of peat in dried blood.
After the close of the meeting the following paper was received by
the secretary, in the transmission of which Mr. Gladding said:
85
The question arises whether a ooinposite method such as I suggest, a combination of
the present official Kjeldahl and Gunning methods, is otticial per se by virtue of the
official character of the two component methods. If not, I wish that such a combina-
tion method could be formally declared official, as I wish to use it in my laboratory on
account of its far greater reliability. Will you kindly see that the proper action is
taken in this matter?
The report is accordingly submitted for the information of the ref-
eree on nitrogen without instructions from the association.
OOMPAEATIVE WOKK ON NITROGEN ESTIMATIONS BY THE KJELDAHL
AND GUNNING METHODS AND BY A COMBINATION OF THE TWO
METHODS.
By Thomas S. Gladding.
A number of fertilizers of all kinds — blood, tankage, fish scrap, humus, etc. — received
in the course of lousiness have been recently examined for nitrogen by either the official
Kjeldahl en' Gunning method and at the same time a comparative analysis has been
made l)y a combination of the two methods as follows:
COMBINATION METHOD.
One gram of fertilizer; 25 cc of sulphuric acid; 10 grams of potassium suxphate; 0.7
gram of mercuric oxid; heat till water white. Cool, add 200 cc of water, 0.5 gram of
zinc dust, 25 cc of potassium sulphid solution, 50 cc of soda solution, and distil.
The results of the 81 comparative analyses given in the following table show an
almost uniformly higher percentage of nitrogen by the combination method. The
average excess is 0.06 to 0.08 per cent. In many individual cases the excess is 0.20
per cent. Our experience shows that the combination method is more rapid and
more reliable.
Comparison of nitrogen determinations made by the Kjeldahl, the Gnnning, and the combi-
nation methods.
Kjeldahl
method.
Combina-
tion
method.
Kjeldahl
method.
Combina-
tion
method.
Gunning
method.
Combina-
tion
method.
■
Per cent.
Per cent.
1 Per cent.
Per cent.
Per cent.
Per cent.
8.56
8.61
4.40
4.42
10.91
11.04
16.11
16.26
15.66
15.71
16.21
16.26
12.06
12.01
11.66
11.66
14.91
15.06
10.57
10.67
2.02
2.02
10.46
10.46
9.61
9.76
6.15
6.18
9.46
9.61
9.06
9.11
6.85
6.90
8.25
8.30
16.51
16.61
9.71
9.81
6.31
6.36
11.66
11.76
4.33
4.40
12.21
12.26
8.35
8.25
7.70
7.80 1
5.70
5.85
7.16
7.06
8.55
8.50 i
7.45
7.40
11.71
11.91
10.01
9.91
11.81
12.06
15.41
15.61
2.85-
2.85
6.23
6.30
15.66
15.71
9.75
9.85
11.46
11.56
8.26
8.31
9.46
9.56
9.31
9.46
11.66
11.71
11.46
11.41
4.10
4.20
11.81
11.91
11.76
11.86
7.81
7.81
7.85
7.90
7.86
7.81
10.26
10.41
9.01
9.01
11.66
11.86
5. 27
5.32
5. 41
5.46
: 11.96
12.11
8.86
8.91
7.06
7.06
1 6.75
6.90 ■
1.50
1.55
10.41
10.36
4.15
4.25
10.56
10.61
8.66
8.66
8.41
8.41
10.96
11.01
11.96
12.06
9.51
9.61
9.46
9.51
7.66
13.71
6.20
11.36
9.31
7.71
13.81
6.25
11.46
9.41
14.61
9.91
1.97
9.36
8.86
8.26
4.30
■ 14. 76
10.01 !
2.02
9.56
9.01
8.36
4.30
:;:::: i
1
::::::::::::::::::::::::
;.... . 1
i
10. 456
10. 515
8.33
8.39
9.107
9.187
EEPOET ON THE SEPARATIOlf OP NITROGENOUS BODIES IN CHEESE.
By R. Harcourt. Referee.
Last year collaborative work was done with the method proposed by Van Slyke
and Hart for the separation of nitrogenous bodies in cheese. « The results obtained
and reported & showed some variations in the amount of nitrogen secured in the
water extract and A'ery wide differences in the quantity obtained from the extraction
with the salt solution. Fmlher study showed that in nearly every case complete
extraction was not secured and that the cotton wool filter did not make a complete
separation of all the insoluble nitrogenous bodies. It is evident that until complete
extractions and filtrations of the soluble nitrogenous substances are obtained there
is little use of doing collaborative work on the separation of the various nitrogenous
bodies contained in the solution. Consequently, it was decided to confine the work
of this year, first, to a study of the amount of water required to make a complete
extraction, and. second, to improving the method of filtering.
Before preparing the following sheet of instructions several methods of filtering and
centrifugal force were experimented with to ascertain the most efficient and rapid
method of seeming a clear filtrate. The use of an asbestos pad, about one-quarter of
an inch thick, on a "Wilt plate or a Hirsch funnel, attached to a strong filter pump,
gave us the best results, and this method was accordingly selected for the cooperative
work.
IXSTRUCTIOX SHEET.
Sampling. — Thoroughly mix the contents of the bottle before taking sample.
Extraction of water-soluble products. — In a porcelain mortar, thoroughly mix 25
grams of cheese sample, prepared as indicated above, with about an equal bulk of clean
quartz sand. Ti-ansfer this mixture to a450-cc Erlenmeyer flask, to which add about
100 cc of distilled water at a temperatm-e of 50° C. The flask is placed on a water bath
or in some place where it can be kept at a temperature of 50° to 55° C, and is allowed
to stand for half an hour, being vigorously shaken from time to time. The liquid
portion is then decanted through a filter of absorbent cotton into a 500-cc flask. The
residue is again treated with 100 cc of water, heated, agitated, and the liquid decanted
as before. This process is repeated until the filtrate, after being cooled to room tem-
peratm-e, amounts to 500 cc, exclusive of the fat, which usually is present at the top
of the liquid.
After the 500 cc of water extract has been obtained, filter through a thick pad of
asbestos on a Wilt plate or Hirsch funnel. Determine the nitrogen by the Kjeldahl
method, in an aliquot portion of the solution, (a) before passing thi-ough the asbestos
pad and (h) after passing thi'ough the a?bestos pad.
Continue the extraction of the insoluble residues with another 500 cc of water in
the same manner as before, and, after filtering through the asbestos pad. determine
the nitrogen in an aliquot portion.
The cotton filter mentioned is made of two la^^ers of absorbent cotton, prepared as
follows: In a glass funnel, place some absorbent cotton to the depth of about 1 inch,
moisten with water, in order to compact it, and then above this place another layer
of cotton of the same thickness. Upon this pour the portions of cheese extract. This
kind of filter allows rapid filtration without the aid of a pump, and is as effective in
every way as paper, which requires half a day or more for complete filtration of 500 cc
of extract. Several samples of cheese can be extracted at the same time. The upper
layer of cotton holds all solid particles and can be returned to the flask for extraction
with salt solution.
Extraction of salt-soluhle products. — Place the upper layer of cotton in the flask with
the residue from the water extraction. Extract with several portions of a 5 per cent
solution of sodium chlorid, the process being carried out both with regard to tempera-
ture and method of filtering the same as in preparing the water extract. Filter through
the asbestos pad and determine the nitrogen by the Kjeldahl method before and after
passing through the asbestos pad.
a U. S. Dept. Agr.. Bureau of Chemistry, Bui. 73. p. 87. Proceedings of 1902.
& U. S. Dept. Agr., Bureau of Chemistry, Bui. 99, p. 125, Proceedings of 1905.
.87
Continue the extraction of the insoluble residue with another 500 cc of the 5 per
cent salt solution, filter through an asbestos pad, and determine nitrogen as before.
Report the duplicate work in nitrogen as percentages of weight of cheese taken.
Eleven chemists signified their willingness to cooperate in the work, and, in June,
two samples of cheese were sent to each one, together with a copy of the above instruc-
tions. For various reasons, only four of those who received samples were able to
make complete returns. The results are as follows:
Results of cooperative tvork on determination of nitrogen in cheese.
Deter iinnation.
Water extract:
First 500 cc through cotton wool
alone
First 500 cc through cotton wool
and asbestos
Second 500 cc through cotton
wool alone
Second 500 cc through cotton
wool and asliestos
Salt extract:
First 500 cc through cotton wool
alone
First 500 cc through cotton wool
and asl)estos
Second 500 cc through cotton
wool alone
Second 500 cc through cotton
wool and asbestos
E.G.deCorioUs,
Guelph,
Ontario.
Per ct.
faO.946
\ .930
f .857
1 .862
f .141
\ .141
( .117
\ .124
/ 1.098
t 1.098
f 1.042
\ 1.047
f .386
1 .381
Per ct.
0.748
.714
.661
.661
.112
.140
.140
.140
1.350
1.355
1.305
1.316
.476
.482
.426
.426
A. W. Bos-
worth,
Geneva. N. Y.
Per ct.
1.07
1.08
1.09
1.09
.18
.18
.15
.14
.59
.59
.54
.52
.32
.33
.33
.33
Per ct.
1.06
1.05
1.07
1.04
.18
.20
.17
1.42
1.40
1.44
.34
.37
.33
.33
A. W. Dox,
Storrs, Conn.
Per ct.
1.16
1.12
1.11
1.11
.10
.10
1.37
1.37
1.34
1.85
Per ct.
0.97
.96
.12
.11
1.83
1.84
1.80
1.81
B. Pilkington,
CorvalUs,
Oreg.
Per ct.
1.101
1.067
.946
.966
.078
.078
.949
.724
.724
.674
.701
.612
.623
Per ct.
1.145
.927
.922
.921
.179
.185
.199
1.438
1.314
1.117
1.117
.517
.494
.589
.613
a The cheese was mailed to the various collaborators during very hot weather. The sample used
by Mr. de Coriolis was kept at a temperature of 40° F. until the work was comm-nc^d. This differ-
ence in temperature and th^ changes it would induce doubtless explain why he recovered a smaller
amount of water-soluble nitrogen. However, this does not destroy the value of results, as the only
points studied were with reference to the completeness of the extraction and filtration.
The above figures clearly show that the 500 cc of water proposed in the original
method were not sufficient to make a complete extraction of the water-soluble nitrog-
enous substances. With comparatively "green" cheese the 500 cc of water may
be sufficient, but when the cheese is well ripened there is a tendency for the material
to gather into a glue-like mass which prevents the extraction of the soluble compounds.
In the case of the salt solution extraction there is so much nitrogen recovered in the
second 500 cc that it is quite possible that even more than 1,000 cc of the solution
should have been used in making the extract.
With reference to the method of filtering, it is evident that two of the workers were
able to make a complete separation of the soluble bodies by means of the cotton wool.
In our own work we have found it convenient to manipulate the cotton filter so as to
cause no delay in the extraction, even if some of the finer particles of insolul^le matter
do pass into the filtrate, and then after the full extraction has been made to draw the
whole filtrate through a thick pad of asbestos on a Wilt plate or Hirsch funnel. The
fat is retained by the cotton and the filtration through the asbestos can be made quickly
and insures a perfectly clear filtrate.
Recommendations.
It is recommended:
(1) That the original method of preparing the water extract, proposed by Van Slyke
and Hart, be so altered as to call for the use of 1,000 cc of water in place of 500 cc.
(2) That the method of drawing the water extract through a thick pad of asbestos,
after it has been separated from the fat and insoluble nitrogenous matter by cotton
wool, be further studied.
(3) That the temperature at which the extraction is made be further studied.
(4) That the completeness of the extraction of matter soluble in salt solution be
further studied.
The report of the referee on the separation of vegetable proteids
was not received until after the association had adjourned, but is
here inserted in its logical sequence.
EEPOET ON THE SEPAEATION OF VEGETABLE PEOTEIDS.
By Harry Snyder, Associate Referee.
In the separation of the alcohol-soluble proteids of wheat and flour a 70 per cent
solution of alcohol is employed. Some analysts prepare the solution by volume and
some by weight. Osborne and Voorhees,^^ in discussing the properties of gliadin,
state: "In absolute alcohol the proteid (extracted by dilute alcohol) is entirely
insoluble, but dissolves on adding water, the solubility increasing with the addition
of water up to a certain point, and then diminishing. TKe exact degree of solubility
has not been determined for various strengths of alcohol, but mixtures of about 70 per
cent of alcohol and 30 per cent of water dissolve the proteids in almost indefinite
amount."
The absolute strength of the solution which they employed does not appear to be
given. The subsequent treatment which the alcohol-soluble proteids received was
such as to prevent the presence of other proteids. In their work it was not the specific
aim to make a quantitative separation of the proteid, but to secure it in as pure a
form as possible for analysis.
Teller found the amounts of nitrogen dissolved by different strengths of alcohol
to be as follows:
Amounts of nitrogen dissolved by varying strengths of alcohol.
Alcohol
by volume.
Alcohol-
soluble
nitrogen.
Alcohol
by volume.
Alcohol-
soluble
nitrogen.
Per cent.
95
90
85
80
75
70
Per cent.
0.21
.31
.61
.89
1.08
1.18
Per cent.
65
60
55
50
45
40
Per cent.
1.30
1.38
1.40
1.40
1.40
1.40
Because of the lack of constancy in the amount of the alcohol-soluble nitrogen
Kjeldahl has seriously doubted the existence of a definite proteid known as gliadin,
and regarded the alcohol-soluble proteid as "spit off" from the general mass of wheat
proteids. Other investigators have expressed a similar view.
An examination of the results available shows that alcohol of 0.90 specific gravity
(65 per cent by volume and 58 per cent by weight) extracts from one to two-tenths of
a per cent more nitrogen than alcohol that is 10 per cent stronger. The maximum
amount appears to be dissolved by alcohol of 40 to 60 per cent strength, but this
alcohol also extracts the maximum of other proteids.
The following amounts of nitrogen were obtained by alcohol of 60 to 72 per cent
strength by weight in the Minnesota Experiment Station Laboratory:
« Amer. Chem. J., 1893, 15: 438.
89
Nitrogen disfiolre'I hy 60-72 per cent alcohol.
Alcohol.
Nitrogen.
Per cent.
Per cent.
60
0.85
68
.74
70
.70
72
.67
Mr. Frank T. Shiitt, chemist of the Department of Agriculture of the Dominion of
Canada, from whom the sampk^ was obtained, reported 0.70 per cent of nitrogen soluble
in 0.70 per cent alcohol by weight.
In a second sample of high grade Minnesota patent flour the following results
were obtained l^y Mr. Hummel, assistant chemist of the Minnesota experiment station:
Nitrogen results obtained under varying conditions of extractioii {Ilummiel).
Alcohol
Nitrogen
Time of
Alcohol
Nitrogen
Time of
hy weight.
dissolved.
extraction.
hy weight.
dissolved.
extraction.
Per cent.
Per cent.
Hours.
Per cent.
Per cent.
Hours.
70
0.95
24
75
0.68
24
70
.97
30
75
.66
44
72
.88
24 ■
81
.33
24
.90
m 1
81
.35
44
Mr. Shutt also reported the following results on another sample of flour:
Gliadin nitrogen dissolved hy varying strengths of alcohol (Shutt).
Solvent
alcohol.
Gliadin
nitrogen.
Solvent
alcohol.
Ghadin
nitrogen.
Per cent
by weight.
62.5
65
70
Per cent.
0.94
.93
* .92
.83
Per cent
hy ic eight.
76.8
86.4
Per cent.
0.60
.66
.12
The analyses given in the table below were transmitted by Mr. Shutt with the
following' comment:
The ''baking strength" was obtained by the cereal division of the Central Experi-
ment Farm. All of the other data are from the chemical laboratory.
I have arranged the flours in three orders, according to (1) total protein, (2) gliadin,
(3) dry gluten. You will observe that we have in these data a very strong indication
if not a proof of the close relationship between these determinations. A year or
two ago we made a number of estimates of gliadin nitrogen, extending the time of
extraction and varying the agitation and tlie solvent to which the flour was subjected.
We did not, however, find any material increase arising from thus prolonging the
extraction or from the extraordinary shaking.
In an earlier letter Mr. Shutt stated that he believed that lack of care in the matter
of the strength of the alcohol employed as a solvent was the cause of the lack of uni-
formity of gliadin results obtained by different workers.
90
Analysis of flours, Aprils 1906 (Shutt).
[70 per cent alcohol.]
Designation.
Albumi-
noids
(NX 5.7).
Gliadin
(NX 5.7).
Albumi-
Gluten.
No.
noids in
the form
of gliadin.
Wet.
Dry.
Ratio
of dry
to wet.
81
Aurora, C. E. F., 1905
Per cent.
10.60
13.33
9.17
11.79
9.40
13.22
15.50
14.39
16.13
14.42
11.51
13.96
Per cent.
4.56
6.49
3.93
5.64
4.21
6.61
8.03
7.18
7.29
6.72
6.04-
6.38
Per cent, l Per ct.
43.02 36.79
48.68 ■ 43.07
42.85 i 28.67
47.82 i 41.45
44.78 1 32.26
50.00 I 44.14
51.80 1 54.22
49.89 1 58.57
45.18 57.63
46.60 49.37
52.47 42. ^"1
Per ct.
12.01
14.09
10.62
14.89
10.97
15.13
17.89
18.70
18.05
17.36
16.17
18.67
3 06
83
Red FifeH., C. E. F., 1905
3.06
85
Soft Red Fife, Man., 1905
2.70
86
Hard Red Fife, Man., 1905
2 79
87
Assiniboia, Man., 1905
2.94
88
90
Huron (selected) C. E. F., I'j05
8 C , C. E F , 1905
2.91
3 03
91
9 J. 3, C. E. F., 1905
3.13
9'^
Turkey Red, fr. Kansas
3.19
95
98
Advance, C. E. F., 1905
Laurel, C. E. F., 1905
2.84
2.63
99
Colorado No. 50 fr. Colorado
45.70
45.62
2.44
No.
Designation.
Physical characteristics.
Baking
strength
(C.E.S.n
81
Aurora, C. E. F., 1905
Good quaUtj'" resilient
91.0
83
Red Fife H., C. E. F., 1905
do
98.5
85
Soft Red Fife, Man., 1905
Hard Red Fife, Man., 1905
Assiniboia, Man., 1905.
Huron (selected) C. E. F., 1905
8C., C. E. F., 1905
9J. 3, C. E. F., 1905
Turkey Red, fr. Kansas
Advance, C. E. F., 1905
Laurel, C. E. F., 1905
Colorado No. 50, fr. Colorado
do ...
89.0
86
do
100.0
87
88
Yellowish; lacking somewhat in cohesiveness. . .
Lacking somewhat in resiliency
91.5
87.5
90
Good, but slightly sticky
95.0
91
79.0
q9.
Good quality resilient
100.0
95
do
91. U
98
do . . .
77.0
99
Good, but rather granular and lacking some-
what in elasticity.
94.0
Recommendation .
There is a close agreement between the results obtained upon the same sample of
flour by the two laboratories reporting, and there is also a uniformity in the results
secured when the strength of the alcohol is varied and different samples of flour used,
i. e., there is a decrease in the alcohol-soluble nitrogen with increase in strength of
solvent.
To secure greater uniformity of results, it is recommended that for the extraction
of alcohol-soluble nitrogen in wheat and flour 70 per cent alcohol by weight, specific
gravity 0.871, be employed.
On behalf of the Washington Section of the American Chemical
Society, an invitation was extended to the association to be present
at the meeting of the section to be held at the Cosmos Club, Novemiber
15, the subject under discussion being denatured alcohol, with a paper
on the new denatured alcohol law and its effects, by Mr. C. A. Cramp-
ton of the Treasury Department.
The association adjourned.
THURSDAY— MORNING SESSION.
At the opening of the morning session Vice-President Street took
the chair and President Hopkins dehvered the annual address, upon
the subject ''The duty of chemistry to agriculture." This address
has been printed as a circular of the Illinois station and is omitted
here.
At the close of the president's address the following committees
were announced :
* APPOINTMENT OF COMMITTEES.
Committee on resolutions: Messrs. Van Slyke, Tolman, and Allen.
Committee on amendments to the constitution: Messrs. Winton, A. M. McGill, and
Veitch.
Committee on nominations: Messrs. Woods, Peter, Woll, Blair, and Frear.
Committee A on recommendations of referees: Messrs. Davidson, Hartwell, McDon-
nell, Magruder, and Penny.
Committee B on recommendations of referees: Messrs. Holland, Kebler, Bartlctt,
Robison, and Woll.
Committee C on recommendations of referees: Messrs. Lythgoe, Bigelow, Doolittlo,
Cochran, and CD. Howard.
Committee on revision of methods: Messrs. Haywood, Veitch, Tolman, Winton,
Street, Woll, and Pettit. (This committee had been appointed before the meeting in
order that a report might be presented.)
Committee to invite the Secretary and Assistant Secretary of Agriculture to address
the association: Messrs. Davidson, Hardin, and Patterson.
EEPOET ON THE SEPAEATION OF MEAT PEOTEIDS.
By F. C. Cook, Associate Referee.
To those who signified a willingness to cooperate, a sample of a commercial meat
extract, together with directions for the work, was sent on July 25, 1906. Cooperators
were asked to keep the sample cold and to commence the work as soon as possible,
conducting it according to the following directions:
DIRECTIONS FOR MEAT EXTRACT ANALYSIS.
Insoluble and coagulahle proteid. — Use 8 grams of the sample in duplicate, dissolved
in 75 cc of distilled water. Make neutral to litmus (test neutrality, taking a drop
outside, by means of a capillary tube, on a piece of delicate litmus paper) by adding
(91)
92
dilute alkali, tenth-norinal preferred. Report nuniber of cubic centimeters of tenth-
normal alkali per 100 grams of sample. Acidifj- the neutral solution by adding 10
cc of tenth-normal acetic acid and make the total volume up to 120 cc. Coagulate
by boiling for two minutes. After "cooling, iilter and -wash with distilled water.
Determine the nitrogen in the precipitate, which is that of the insoluble and coagu-
lated nitrogen. The filtrate is made up to 150 cc volume and divided into three 50 cc
aliquots.
Albumosts at room temperature. — ^Transfer one 50 cc aliquot to a Kjeldahl flask,
add three drops of sulphuric acid, and saturate with zinc sulphate, shaking fre-
quently. Have 1 gram of zinc sidphate undissolved in the bottom of the flask. Let
stand over night, filter and wash with saturated zinc sulphate solution after having
rinsed the flask out with a portion of the saturated zinc-sulphate solution. Eeturn
the filter paper to the flask and determine the nitrogen, which is that of the albumoses.
Alhumosts at 7<J^ C. — ^Use a 50 cc aliquot of the coagulated nitrogen filtrate and
proceed as in the case of proteoses in the cold. After completely saturating with zinc
sulphate heat at 70° C. for 10 minutes, allow to stand overnight, and proceed as above.
Peptones, tannin and salt method. — Transfer a 50 cc aliquot of the coagulated nitro-
gen filtrate to a 100 cc fliisk. Add 15 grams of sodium chlorid and shake to dissolve
as much of the salt as possible. Place the flask in the ice box at 15° C. A flask of
distilled water and a second flask containing the tannin solution to be used are simul-
taneously placed in the ice box. The tannin solution is made by dissolving 24 gi-ams
of pure tannic acid in 76 cc of water and filtering. After standing for one hour in
the ice box, 30 cc of the cold 24 per cent tannic-acid solution are added, then cold
distilled water up to the mark, and the contents of the fl^sk well mixed. After stand-
ing in the ice box over night the solution is filtered into a 50 cc flask at ice-box tem-
perature and the nitrogen determined in the 50 cc of the filtrate. As the tannin
used may contain nitrogen it is ad^-isable to nin a blank. Determine the nitrogen
in 50 cc of the blank filtrate. The nitrogen in the 50 cc of the tannin-salt filtrate,
after subtracting that found in the 50 cc filtrate of the blank determination, multiplied
by two gives the amido nitrogen. The total nitrogen minus (amido nitrogen plus
insoluble and coagulated nitrogen) gives the albumose and peptone nitrogen. The
peptone nitrogen is obtained by deducting from the above figure the amount of nitro-
gen in the zinc-sulphate precipitate.
Report results as follows, using the amoimt of nitrogen found by the albumose
method at room temperature when calcidating the peptone nitrogen:
Per cent.
Total nitrogen
Insoluble and coagulable nitrogen
Albumose nitrogen
Albumose nitrogen at 70° C
Peptone nitrogen
Amido nitrogen
The restdts reported by the co<:)peratLng chemists are given in the following table:
Xitrogen determmations on sample of commercial meat extract.
>.itTogen.
AnalTsT.
Aciditv.
Insoluble
and coag-
ulable.
Albumose
at 70° C.
Albumose
at room
tempera-
ture.
Peptone. Amido.
cc tenth-nor-
mal alkali
perlOO grams.
Leavenworth ; 751.0
Jackson 641.0
Cook 625.0 [
Hanford 812.0
Stookev 727.0 '
Per cent. Per cent. Per cent. Per cent.
0.080 2.14 2.18 2.88
.0852 ^ .799 .72S 6.896
.203 1 2.18 , 2.10 3.06
.168 1.15 1.83 1.60 1
.24 1.39 : 1.40 1.10
Per cent.
2.12
.oo
2.90
4.66
5.53
Co:mmext bt the Refekee.
The results show a fair degree of uniformity, considering that variations were to be
expected from the fact that a new method was being tried ( the modified tannin-salt
method). Furthermore, the proteid separation methods are of necessity far from
93
satisfactory, due to our ignorance of the nature of the proteid niolecuh> and the close
relationship of its various hydrolytic products. The work is especially valual)le for
the criticism which it evoked and the comments l)y those cooperating and also by some
who sent in no analytical results.
The acidity determination as carried out ])y titration with tenth-normal alkali,
using litmus as an indicator, has met with much criticism, the figures varying from 625
to 812 cc.
The referee is anxious to obtain a more satisfactory method, but the use of litmus
seems to be the best procedure at present, as the end point with phenolphthalein is
masked by the dark color of these solutions.
The insoluble and coagulable nitrogen figures show great variation, and it is probaV)ly
advisable to use a definite volume of wash water, as incomplete washing gives high
results and long-continued washing yields low results. The two determinations
should be separated, as the coagulable proteid figure is an important one and should
be given alone.
The results obtained with zinc sulphate at room temperature are fully as high as
those obtained by heating to 70° C. for ten minutes, the duplicates agree better, and
the process is simpler.
The tannin-salt method for the separation of proteoses and peptones from simpler
amido bodies is a method not likely to yield good results on first trial. The solution
employed for this determination was undoubtedly too concentrated, and better results
would have been obtained with a more dilute solution. The method will be further
tested by various chemists, and with more careful directions good results can undoubt-
edly be obtained.
In using the tannin-salt method the solution foams to an objectionable degree during
the Kjeldahl digestion unless precautions are taken to prevent it. The procedure
followed by T. C. Trescot, of the Bureau of Chemistry, is to transfer 50 cc of the tannin-
salt filtrate to a Kjeldahl digestion flask and add a few drops of sulphuric acid. Place
the flask in the steam bath, connect with the vacuum, and evaporate to dryness. In
the digestion process about 30 cc of sulphuric acid are added but no potassium sul-
phate. The large amount of sodiimi chlorid used in the process forms sufficient
sodium sulphate, which acts in the same way as the potassium sulphate. The
remainder of the process is carried out in the usual manner, and with the above modifi-
cations there is no diflftculty in using the tannin-salt method in connection with the
Kjeldahl digestion.
Comments by Analysts.
H. S. Grindley, of Urbana, 111., says: "The method suggested for the determination
of the acidity is very unsatisfactory. The determination of the insoluble and coagula-
ble proteid in such a sample we find difficult because of the slow filtering. I think,
also, that the quantity of sample taken for the determination of the secondary proteid s
by tannin salt is too large to give accurate results."
H. C. Bradley, of Madison, Wis., criticises the method as tedious. The litmus-
acidity method is unsatisfactory, and he suggests that the acidity be determined in a
separate sample, using a very dilute solution and phenolphthalein as indicator. He
asks if in the determination of the various kinds of nitrogen present there is any
significant gain made. Mr. Bradley suggests that we determine the more definite
constituents of muscle, namely, creatin, creatinin, and xanthin bases.
E. E. Smith, of New York City, writes that the neutral point as obtained with litmus
paper is inexact and that slow filtration of the insoluble and coagulable proteid is due^
in part to excessive acidity and in part to the attempt to filter cold. The 10 cc of
tenth-normal acid added he considers excessive, as there is so little proteid present.
Half saturating with sodium chlorid, Mr. Smith says, would facilitate coagulation
and preserve the sample. In regard to coagulation he claims that usually the best
94.
results are obtained by the addition of some sodium chlorid, the application of heat
to the boiling point while in a neutral condition, the addition of just a sufficient
amount (which is variable) of acetic acid to cause the separation of the albuminous
substances in a flocculent form, maintaining in a hot condition on the water bath for
the complete and flocculent separation of the albuminous substances, and finally the
filtration while hot and washing with hot water.
L. B. Mendel, of New Haven, Conn., writes that acidity might well be determined
on a separate sample. In regard to the albumose determination and the washing with
saturated zinc-sulphate solution Mr. Mendel suggests that the bumping in the Kjel-
dahl determination, due to zinc sulphate crystallizing out on the filter papers, might
be avoided by washing with a zinc-sulphate solution not concentrated sufficiently to
deposit crystals by surface evaporation. He suggests that the creatin and purin con-
tent be especially considered, as these are the factors which characterize the product
as a meat product.
L. B. Stookey, of Los Angeles, Cal., considers the methods as a whole to be quite
satisfactory, but thinks the quantities used too large to giv,e the best results.
Holmes C. Jackson, of Albany, N. Y., suggests that we discard the litmus method
and employ a physical method for the acidity test, or titrate in dilute solution, using
phenolphthalein as indicator, and add potassium oxalate to counteract precipitated
phosphates or ammoniacal disturbances. Mr. Jackson suggests that for more accurate
work we reprecipitate the proteoses out of a moderately concentrated solution. He
also considers that the supposition that 50 cc of the filtrate from the tannin-salt precip-
itate represents one-half of the nonprecipitable nitrogen is unwarranted in view of
the heavy precipitate present, and instead of being one-half it is more nearly five-
eights of the liquid; the solution employed for the tannin-salt treatment is too con-
centrated and should be diluted.
Results Obtained by the Referee.
The following studies include work on coagulable proteids, proteoses, and the sepa-
ration of organic and inorganic phosphorus.
COAGULABLE PROTEIDS.
A cold-water extraction of beef was made, the sample being filtered and neutralized
to litmus before it was divided into aliquots. All volumes are of 70 cc.
Table 1. — Amount of nitrogen precipitated from cold-water meat extract by varying
amounts of acetic acid and time of boiling.
Time of boiling.
Acetic acid added (tenth-normal) .
5cc.
10 cc.
15 cc. 20 cc.
Gram.
0.0011
Gram.
0. 00084
j
Gram. Gram.
0. 00084 0. 00112
Two minutes
.00168 1 .00112
. 00084 . 00084
Four minutes
. 00168
.00112 .00112
Two pounds of lean beef were heated at 50° C. for two hours, ground and pressed,
diluted, filtered, neutralized to litmus, refiltered, and aliquots taken. In cases where
more than 5 cc of tenth-normal acid were added (10, 15, and 20 cc) the solutions did
not coagulate thoroughly and filtration was impossible. With the addition of 5 cc of
tenth-normal acetic acid the following results were obtained: Just boiled, 0.0578
gram; boiled two minutes, 0.0575 gram; boiled four minutes, 0.0589 gram.
A cold-water extraction of another sample of meat was prepared and neutralized
to litmus before dividing into aliquots as before, the volumes, however, being made
up to 60 cc. The following results were obtained:
95
'able 2. — Nitrogen 'precipitated from a cold-jvater extract of meat under varying con-
ditions.
Time of i)()iliiig.
Neutral
reaction.
Tenth-normal
acetic acid.
2cc.
10 cc.
Just boiled
Gram.
0.0168
.0146
.0156
Gram.
0. 0236
. 0246
.0240
Gram.
0. 0260
. 0261
Four roinutos
.0293
It is evident from these data that neutral solutions are not favorable to coagulation,
as is well known, and that an excess of acid prevents coagulation. The time of boiling
affects the results and it appears that the boiling should be continued for two minutes
at least in order to insure complete coagulation.
A cold-water extraction of quail was used for the next study. Thirty grams of quail
meat were shaken with cold water, washed until the filtrate was clear, and made up
to 500 cc after filtering. All aliquots were made up to 70 cc with water. Two 50 cc
aliquota of the original solution were diluted to 70 cc with water and coagulated by
boiling two minutes. The acidity was due to the natural acid of the fresh meat juice,
amounting to 1.45 cc of tenth-normal alkali per gram or 4.35 cc tenth-normal acid
per 50 cc aliquot. The results are given as No. 9 and No. 10. The remaining 400 cc
of the original solution were neutralized to phenol phthalein, made up to 440 cc vol-
ume and divided into eight 55 cc aliquots. All solutions were boiled two minutes.
Table 3. — Nitrogen precipitated from cold-water extract of quail, using varying amounts
of acetic add {filtered before neutralizing).
Sample.
Tenth-
normal
Nitrogen
precipi-
Description of
filtrate.
acetic acid.
tated.
Cc.
Gram.
1
2
0. 0146
Faintly turbid.
2
2
.0148
Do.
3
5
.0149
Clear.
4
5
.0149
Do.
5
10
.0079
Turbid.
6
10
.0079
Do
7
15
.0039
Faintly turbid.
8
15
.0045
Do.
9
. 0143
Slightly cloudy.
Do.
10
.0143
Four hundred cubic centimeters of a cold-water extract of quail meat, represent-
ing 30 grams of meat, were neutralized, filtered, and made up to 500 cc volume. This
was divided into ten 50 cc aliquots. All solutions were boiled two minutes and the
following results obtained.
96
Table 4. — Nitrogen precipitated from cold-water extract of quail, using varying amounts
of acetic acid {neutralized before filtering) .
Sample.
Tenth-
normal
Nitrogen
precipi-
Description of
filtrate.
acetic acid.
tated.
Cc.
Gram.
1
0
0. 0146
Turbid.
2
0
.0118
Do.
3
5
. 0202
Faintly tiirbid.
4
0
.0194
Do.
5
10
. 0205
Clear.
6
10
. 0210
Do.
7
15
.0210
Do.
8
15
.0199
Do.
Solutions dis-
9
10
20
20
carded as tur-
1 bid and fil-
tered slowly.
In Table 4 the quail juice used was neutralized before filtering and consequently
the acid albumin in this case was removed. In Table 3 the filtration was carried
out before neutralizing and the acid albumin consequently remained in the filtrate.
The results in Table 4 are higher than those in Table 3, and as each solution in both
Tables 3 and 4 represents the same amount of soluble nitrogenous bodies the figures
are interesting. It appears that after the acid albumin has been removed the added
acid is left free to act with the salts in solution and the heat employed on the proteid
matter which is in the coagulable form. In the cases of No. 9 and No. 10 of Table 3,
where the natural acidity of the meat is unaltered, lower results are obtained than
any of Table 4 except No. 2, in which case some error is evident. The amount of
salts present is the same in all cases, and the results can not be ascribed to their influ-
ence. In Table 3, where 5 cc of acetic acid were added, the results are best, as a
clear filtrate is indispensable. In Table 4, 5 cc of acid yield a turbid filtrate and 10
cc are necessary to give a clear filtrate, while 10-cc in Table 3 give a turbid filtrate.
That solutions of muscle proteids are rendered acid by coagulation has been thor-
oughly demonstrated by G. N. Stewart. « A factor secondary to the reaction is the
salt content of the solution. Many workers have demonstrated that an albumin solu-
tion, whose salts have been removed by dialysis, will not coagulate, but on adding acid
a soluble acid albumin may be formed, which on continued boiling is likely to be
changed to an albumose. A good plan seems to be the one suggested by Cohnheim.^
and recommended by E. E. Smith in this report, namely, the addition of sodium
chlorid or some other neutral salt to the neutral solution to be coagulated and then
a varying amount of acetic acid. It is essential to keep the reaction of the solution
as slightly acid as possible, or to add a large amount of salt to insure complete coagu-
lation and to hinder the formation of acid or alkali albumins. It seems doubtful if
any universal rule for the coagulation of proteids can be framed. The proteids
obtained from different sources are so diverse, and animal proteids themselves
differ so much, that it seems necessary to treat each sample individually and so
adjust the conditions to each sample as to give the maximum coagulation and a
clear filtrate.
PROTEOSE RESULTS AT ROOM TEMPERATURE AXD AT 70° C.
In all cases a neutralized filtered aliquot of the filtrate obtained from the determina-
tion of coagulable proteids was employed in this study. To this were added 3 drops of
sulphuric acid, and the solution was then saturated with zinc sulphate. The deter-
minations at 70° C. were saturated as is the usual custom and the saturated solution
oj. Physiol., 1899, ^4.- 450.
b Zts. physiol. Chem., 1901, SS: 455.
97
heated to 70° C. for ten minutes, then allowed to stand overnight. Solutions of meat
extracts and water-soluble portions of beef and quail were used in making the following
determinations:
Table 5. — Comparison of proteose results at room temperature and at 70° C.
Serial No.
Nitrogen precipitated.
Serial No.
Nitrogen precipitated.
At room
tempera-
ture.
At 70° C.
At room
tempera-
ture.
At 70° C.
1813
18372
18634
15416
18609
Gram..
0.0272]
.07916
.08309
.1124
.Q160
Gram.
0. 0275
.08562
.08224
.1075
.0138
17072
17599 (a)
17599 (6)
17C00 (a)
17600 (ft)
Gram.
0.0994
} .00168
1 .00116
Gram.
0. 0982
j .00196
\ .00168
/ .00116
\ .00087
The above results are indecisive as to which method gives higher results. In all
cases the differences are slight, and as the room temperature method gives better
duplicates and is simpler it seems desirable to follow that procedure.
SEPARATION OF ORGANIC AND INORGANIC PHOSPHORUS.
Two methods were applied to meat extracts for separating these two forms of phos-
phorus, namely, the Siegfried-Singewald « method and the modified method of Hart
and Andrews.^ The latter could not be successfully applied to meat extracts, as
the molybdic acid caused a heavy precipitate of the proteid bodies.
Organic phosphorus determinations by the Siegfried-Singewald method,
[Grams per 10 cc]
T^tal
phosphoric
acid.
Organic
phosphoric
acid.
Organic
phosphoric
acid in
presence
of added
sodium
phosphate.
0.0117
.0140
0.0017
.0017
0.0023
.0016
To another sample 40 cc of a lecithin solution, containing 0.0005 gram of phosphoric
acid per 10 cc were added, and the Siegfried-Singewald method applied. After
deducting the amount of organic phosphorus which was added, the results show 0.0035
gram of phosphoric acid in the organic form per 10 cc, as compared with 0.0017 gram
when no phosphorus was added. From the few results given above and the application
of this method to some thirty samples of Commercial meat extracts, it is the referee's
opinion that the method does not effect a separation of the organic from the inorganic
phosphorus.
Siegfried's method was also compared with the ordinary ether and alcohol extraction
method for organic phosphorus. By the former method 0.16 per cent of phosphoric
acid was found in organic combination in a solution of meat extract, and by the
latter method 0.08 per cent. In all cases total phosphoric -acid was determined by
Neumann's c method. It is intended to investigate during the coming year the ether
and alcohol extraction method for organic phosphorus determination.
«Zts. Nahr. Genussm., 1905, 10: 522.
bAmer. Chem. J., 1903, 30: -182.
^Zts. physiol. Chem., 1902, 37: 115.
31104— No. 105—07 7
98
Conclusions and Outline op Work.
It seems advisable to give the question of the acidity determination in meat extracts
and related products further study and to ask for cooperation on this point another
year. The coagulable proteids should be determined separately from the insoluble
proteids. The best conditions for the determination of coagulable proteids will be
further studied. It appears that the individual factor is so great that it may be impos-
sible to formulate a general method applicable to all meat products. The amount of
water used in washing the insoluble proteid matter should be definitely stated.
The method of determining albumoses by heating to 70° C. for ten minutes after
saturating with zinc sulphate gives no higher results than the method of saturating
with zinc sulphate at room temperature. The room temperature method is simpler
and gives better duplicates.
No satisfactory conclusion can be reached in regard to the application of the tannin-
salt method, as herein outlined, from the limited number of cooperative results
obtained. With more complete directions and the use of a more dilute solution better
results might have been obtained by the various collaborators. The method needs
further study and trial in the hands of experienced workers. The creatinin determi-
nation, as well as the question of organic phosphorus separation, should be taken up by
the referee in 1907.
EEPOET ON DAIET PEODUOTS.
By F. W. WoLL, Referee.
The referee was requested, by vote of the association last year, to continue the
study of the preservation of milk samples for the determination of milk proteids,
and to take up two new subjects for study, namely, methods of determining sugar
in condensed milk, dried milks, or milk powders, and, second, methods of detecting
the adulteration of butter with small quantities of foreign fats.
In view of the fact that the first subject mentioned can only be satisfactorily worked
out by individual effort, as is indicated by the small interest taken in the subject by
our chemists in the past, and furthermore because experience has shown that the
cooperative work done stands in nearly inverse ratio to the amount outlined for study,
your referee decided to recjuest the cooperation of chemists in a study of only two
subjects:
(1) The analysis of condensed milk, including determinations of sugar and fat, and
(2) The Gottlieb method of determining fat in skim milk and other dairy products.
By arrangement with the associate referee on adulteration of dairy products, Mr.
Leach, the third subject recommended by the association for cooperative work,
methods of detecting adulteration of butter with small amounts of foreign fats, was
left to be studied by him, if practicable, as properly coming under his domain.
Sixteen different chemists signified their willingness to cooperate in the work
outlined by the referee, and two carefully prepared samples of condensed milk,
sweetened and unsweetened, were furnished them, the following explanatory letter
having been previously mailed:
Madison, Wis., March 27 , 1906 .
Dear Sir: The cooperative work on dairy products for the Association of Official
Agricultural Chemists during this year was outlined in my circular letter of the 3d inst.
With reference to I a (Effect of preservatives upon the determination of albumin
in milk), I would say that I was not present at the last two conventions of the asso-
ciation and do not possess sufficient information in regard to the results of the work
so far accomplished to outline further work on this subject. Chemists who are so
situated as to be able to continue this work are urged to do so; when the proceedings
of last year's convention are published, it is likely that the direction will be shown
in which further study of the subject should go.
99
As to I ]) and II b (Methods of dctormining sugar and fat in condensed milk),
samples of unsweetened cond(>nsed milk will be sent about April 15 to all chemists
who have signified their intention to cooperate in this work. You are requested
to make the following determinations:
Sugar. — In the sample of sweetened condensed milk, (1) gravimetric or volumetric
method with Fehling solution, and (2) polariscope method, both before and after
inversion with citric acid or yeast; direct gravimetric and polariscope methods in the
sample of unsweetened condensed milk.
Fat. — (1) The Babcock asbestos method. -Five grams of a 40 per cent solution of
condensed milk in water are weighed into a copper asbestos tube and after drying
extracted for five hours, as in case of cow's milk. On completed extraction, the
tubes are placed in distilled water to dissolve the sugar, then dried and extracted
for another five hours, when the fat is dried and weighed.
(2) The modification of the Babcock centrifugal method by Leach (Food Inspec-
tion, p. 149) or by Farrington (Wisconsin Seventeenth Annual Report 1900, p, 86)
may be used in the case of sweetened condensed milk and the direct Babcock cen-
trifugal method in case of the unsweetened milk.
(3) The Gottlieb method (see below).
It is requested that, if practicable, comparative analyses by these methods of other
samples of condensed milk and of dried milk or milk powders be made and reported.
As to II a (Comparisons of the official and the Gottlieb methods for the determina-
tion of fat in skim milk and buttermilk), chemists are asked to make comparative
fat determinations in a number of samples of skim milk and buttermilk by these two
methods.
The Gottlieb method was originally described in Landw. Versuchs.- Stat., 1892, 40:
1-27. It is "now the official method in Sweden and Denmark for the determination
of fat in skim milk and buttermilk, recent work by European chemists having shown
that the ordinary extraction methods give too low results for milks low in fat. The
Gottlieb method is as follows:
"Ten cc of milk are measured into a glass cylinder three-fourths inch in diameter
and about 14 inches long (a 100 cc burette or a eudiometer tube will do); 1 cc con-
centrated ammonia is added and mixed thoroughly with the milk; the following
chemicals are next added in the order given: 10 cc of 92 per cent alcohol, 25 cc of
washed ether and 25 cc of petroleum ether (boiling point below 80° C), the cylinder
being closed with a moistened cork stopper and the contents shaken several times
after the addition of each. The cylinder is then left standing for six hours or more.
The clear fat solution is next pipetted off into a small weighed flask by means of a
siphon drawn to a fine point (see fig. 6, Landw. Versuchs. -Stat., 40: 6), which is low-
ered into the fat solution to within 0.5 cm of the turbid bottom layer. After evapo-
rating the ether solution in a hood the flasks are dried in a steam oven for two to three
hours and weighed. This method is applicable to new milk, skim milk, buttermilk,
whey, cream, cheese, condensed milk, and milk powder, but has been found of special
value for determining fat in skim milk, buttermilk, cheese, and condensed milk. In
the case of products high in fat a second treatment with 10 cc each of ether and petro-
leum ether is advisable in order to recover the last trace of fat."
It is requested that reports of the work done along the preceding lines of study be
sent to the referee as soon as finished, so as to render possible further investigation
of points that may suggest themselves. All reports should be in by October 1. I
shall be pleased to correspond with chemists who need additional information in
regard to the work outlined above.
Hoping that you may be able to contribute in some measure to the further advance-
ment of our knowledge of the subjects scheduled for study, I am,
Very respectfully, yours,
F. W. WOLL,
Referee on Dairy Products.
Reports of the work done on these samples we^e received from seven chemists,
located at four different stations, o the results being given in the following pages. For
the sake of convenience the analyses and the discussion of the results are arranged
under four headings: (1) Sugar in condensed milk. (2) Fat in condensed milk. (3)
Fat in dried milk and milk powders. (4) The Gottlieb method of fat determination.
« Results obtained by the Dairy Laboratory, Bureau of Chemistry, U. S. Depart-
ment of Agriculture, were received after this report was written and have been included
in the tables given below.
100
(1) Sugar in condensed milk. — The following results were obtained for sugar in con-
densed milk by the different methods:
Percentage of sugar iyi condensed milk hy different methods — referee's samples.
Analvst.
Gra Timet ric
method.
Polariscopic
method.
Lactose, i Sucrose, i Lactose. I Sucrose.
Sample A, sweetened condensed milk:
Fulton, Massachusetts
Norton. Arkansas
Bartlett . Maine
Jaffa and Stewart, California
Olson. Wisconsin
Sample B. unsweetened condensed milk:
Fulton, Massachusetts
Norton. Arkansas
Bartlett, Maine
Olson, Wisconsin
Patrick and Boyle, Washington, D. C.
13.37
9.5
14 94 I
14 37 '
14 69 j
10.49 I.
9.4 i
10.08 .
n.05 .
10.04 .
40. 62 ! .
45.48
40. 18 _
41. 78 10. 88
35.25
40.90
3S. 67
1L3
11.56
10.07
Other determinations were made on one sample of sweetened condensed milk and
four samples of unsweetened condensed milk at the California and "Wisconsin Experi-
ment stations, as shown below."
Other determinations of sugar in condensed milk.
Gravimetric.
Polariscope.
Lactose. Sucrose. Lactose. Sucrose.
Sweetened condensed milk:
Olson. Wisconsin
Unsweetened condensed milk:
Jaffa and Stewart, California
Olson, Wisconsin
Jaffa and Stewart, California
Per cent. Per cent. Per cent. Per cent.
16.41
35.91
34 25
9.28 9.36
9.91 ! ....J 9.19
8.11 &00
10.87 10.74
9.96 9.84
REMARKS BY ANALYSTS.
Massachusetts. — Solutions were prepared by diluting 200 grams to 500 cc. It is
impossible to prepare sample B satisfactorily because of the churned condition of the
fat and of the action of the formalin, which was present in considerable quantity.
Having no polariscope in condition, the sugar results were obtained by gravimetric
methods.
Maine. — The solution of condensed milk was boiled with the copper solution for
two minutes.
California. — The discrepancy between the determinations of lactose by the polari-
scope and by the Fehling solution I can only account for on the ground that some of the
cane sugar had been inverted or reduced at the time of manufacture. All of the deter-
minations were made in duplicate. The analysis of sample A calculated to normal
milk by use of either the "solids-not-fat'' method, or on the basis of 0.70 per cent ash,
gives a possible milk, if we use the polarimetric determination of lactose of 10.88 per
cent, as follows: Moisture 86.67 per cent, protein 3.38 per cent, fat 4.37 per cent, sugar
4.88 per cent, and ash 0.70 per cent. The results are not satisfactory', however, if we
use the figure 14.78 per cent. The determinations of lactose by the graA-imetric method
were made by weighing the cuprous oxid.
Wisconsin. — Twenty cubic centimeters of a 40 per cent solution of condensed milk
were treated with 0.25 gram of citric acid, made up to 200 cc, filtered, and 20 cc of the
See also report of Messrs. G. E. Patrick and M. Boyle, p. 106.
101
iiltrate were added to me boiling Fehling solution and boiled for six minutes. The
cuprous oxid was collected and washed en the filter, and weighed after reduction to
cupric oxid. After correction for volume and the amount of copper in CuO, the
equivalent in lactose was obtained from the table in Bulletin 46 of the Bureau of
Chemistry and the per cent of lactose in the condensed milk calculated.
For the sweetened condensed milk 100 cc of the filtrate were treated with 1 gram of
citric acid for inversion and neutralized with sodium carbonate, then boiled to pre-
cipitate the albumen, hltered, and made up to 100 cc. Ten cubic centimeters of the
filtrate were taken for quantitative determinations, and boiled with the Fehling solu-
tion for two minutes.
For the polariscopic method 20. 048 grams were treated with 0.5 gram of citric acid,
neutralized with sodium carbonate, boiled ten minutes, made up to 200 cc, and filtered.
The filtrate was polarized in 400 mm and 200 mm tubes. One hundred cubic centi-
meters of filtrate were inverted with 1 gram of citric acid, neutralized, cooled, and
potassium mercuric iodid added, made up to 100 cc and filtered. This solution was
polarized and the amount of sucrose after volume correction was calculated by the
usual formula.
Of the samples of unsweetened condensed milk, 20.493 grams were treated with 0.5
gram of citric acid and neutralized, boiled for ten minutes, cooled, made up to 200 cc,
and filtered. The filtrate was polarized in 400-mm tubes. After correction for volume
was made the per cent of lactose in the milk was obtained.
DISCUSSION OF RESULTS.
The results for lactose in both sweetened and unsweetened condensed milk are
fairly concordant, with one exception in each case. This can not, however, be said
to be the case with the determinations of sucrose in the sweetened condensed milk.
It is a question, however, how much the personal factor enters into the results. It
has often been found in the past that unfamiliarity with the methods under investiga-
tion has been responsible for strange results on cooperative work. Detailed direbtions
as to the manner of procedure were not sent out by the referee in the case of the con-
densed milk work, as it was felt that this year's work was only preliminary, and it was
desired to obtain information as to the present methods of determining sugar in con-
densed milk in vogue among station and food chemists and the kind of results obtained.
It is planned to continue this study next year, in the expectation that as satisfactory
methods as possible for the determination of both lactose and sucrose in condensed
milk may be worked out and adopted as official by the association.
(2) Fat in Condensed Milk.
Fat was determined in the condensed milk by seven analysts, located at five differ-
ent stations, the Babcock asbestos method, the Babcock centrifugal method as modified
by Leach or Farrington, or the Gottlieb method being used by the various chemists.
102
The following table shows the results obtained with these methods:
Percentage of fat in condensed-milk samples.
Analvst.
Asbestos metliod.
Babeocktest.
Gorclieb
Extrac- Extrac- ^^ , i^qrh barring- method,
tionl. tioii2. ^^^^^- ^^^(^^- ton.
Referee's sample A, sweetened condensed
milk:
Smith. Massachusetts
WhiTtier. Massachusetts
2.25
2.43
6.09
5.82
S.34
S.25
7.97
Fulton. Massachusetts
, 9.0
' &4
ao8
"Vorton. -Arkansas
a 4. 70
Bartlett. Maine
Olson. "Vrisconsin.
JafFn jind .^fpwjjrt, Cilifomia.
a is"
6.S4
6. S6
; 5"68"
.45
.42
1 8.76
S.26
9.92
7.29
7.2S
9.'68"
9.75 -.
&7
8.8
S.S1
S.13
Referee's sample B. unsweetened con-
densed miTk-:
"V6.'4S"
fe&9
6 5u93
7.7
fee. 8
7.12
"^Tiittier. Massachusetts
"Vorton A rka n ■m ~!
a 4. 9
Bartlett . Maine
Olson, Wisconsin
7.20
.02
7.34
7.22
^7.55
1 &04
i 9.65
9.70
7.32
! 8.68
7.20
] 1
; 6.90 |{
1 7.03
} '■'^
a 7. 24
Other samples of sweetened condensed
milt:
Olson. Wisconsin
r 5. 7S
\ 7.56
f 7.24
1 &08
2.26
2.09
: .08
i :^-
S.37
9-66
Jaffa and Stewart, California
Other samples of unsweetened condensed
milk:
Olson. Wisconsin
y.30
' ■■ 1 i
7.53
i S.93
Jajla and Stewart. California
6.76 1 !
1 .
1
Averages of comparative determination.^
Method.
Sweetened.
Unsweet-
ened.
Per cent. Per cent.
By extraction method ^ f^ S. til ^ 7. 55
Gottheb method ^ S. 59 '^ 7. 54
Bv extraction method f 9. 29 , / 7. 21.
Babcock test, Leach modification « 9. 3S 7i6. 80
By extraction method /8.51 ! «7.28
Babeocktest (original) / P8.75 j «6.40
a Extracted only once.
t' Babcock original method.
(■ Adams method. See p. 106.
d Fire determinations.
f Three determinations.
/ Two determinations.
g Farrington modification.
REMARKS BY AXALTSTS.
Massachusetts. — In using the Babcock asbestos method it would seem advisable to
continue the fii-st extraction over night, a^ in the case of feedstuffs. then proceed as
stated.
In the centrifugal method the decanting and pipetting processes, as outlined for
the modified Babcock methods, did not prove entirely satisfactory^, therefore the solu-
tions were decanted through parchment filter paper (to prevent loss) and the detached
particles returned to the test bottles. Fifteen cubic centimeters of the diluted sam-
ples were used in the several tests. The Leach modification gave slightly higher
results than the Farrington on the sweetened product. The hardening action of the
formaldehyde was ver^- apparent in the unsweetened sample. Comparatively few
of the sepai^ations could be considered perfect tests.
The Gottlieb method seems rather more applicable to materials of a low fat content
and with such samples blank determinations are evidently necessary. Closed cylin-
drical separator^- funnels should facilitate the process materially.
103
Maine. — Could not make the Babcock centrifugal method come up to the others
on sample B. In the asbestos method two extractions were made, but the fat was
not weighed after the first extractions were completed. The sweetened sample was
extracted, then soaked in water as directed, dried, and extracted again. The unsweet-
ened sample was not treated with water.
California. — The extractions of fat were made in S. & S. thimbles.
Wisconsin. — After the first extractions in the Babcock asbestos method, the cylin-
ders were allowed to soak in a liter of distilled water for about three hours. They
were then rinsed with distilled water and dried in a steam oven at 92° C. for five
hours before the second extraction was commenced. The first extraction was con-
tinued overnight and the second for about eight holirs.
For simplicity and accuracy the Gottlieb method, in our experience, is to be pre-
ferred to the extraction method, especially for skim milk, buttermilk, and milks low
in fat. It also appears to give most satisfactory results for fat in condensed milk and
cheese.
DISCUSSION OF RESULTS.
While the results obtained by the various analysts are not always as concordant as
might be desired, some facts are nevertheless brought out in a striking manner.
First, the necessity of a double extraction in the ether-extraction method. In the
sweetened condensed milk two-thirds or more of the fat may escape extraction when
extracted only once, even if the process be continued for twelve hours or more. In
unsweetened condensed milk the error introduced by making a single extraction
only is, as a rule, small, but even here the safer way is to make a second extraction,
after treating the extraction tubes with water and drying. If this modification is
introduced in the asbestos method, it may safely be made official for fat in condensed
milk. The Babcock centrifugal method, modified by Leach, which is the only
method now given in Bulletin 65 of the Bureau of Chemistry, is provisional only,
and may be depended upon to give satisfactory and correct results when the direc-
tions are followed in detail, but it would seem that with it, as the official method of
the association, the gravimetric ether-extraction method should be given, modified to
provide for double extraction with intervening removal of the sugar so as to insure
complete recovery of the entire fat content of the condensed milk.
. Either of the modifications of Babcock's centrifugal method appears to give satis-
factory results, there being but slight differences in the averages of several determi-
nations by the double-extraction method and the Leach or Farrington modifications.
The average results obtained by the Gottlieb method and by double extraction are
almost identical . In the former method the solution should be treated a second time
with 10 cc of sulphuric and petrolic ether, in the case of condensed milk, as with all
other dairy products containing more than a fraction of a per cent of fat, to allow for
the error of not pipetting off the entire volume of the fat solution, and for the fat adher-
ing to the side of the burette on drawing off the fat solution. The original Babcock
test is not applicable even for unsweetetied condensed milk, as the results, if they can-
be read at all, are likely to be about 1 per cent or more too low (average of three differ-
ent determinations, by double extraction, 7.28 per cent; by original Babcock test,
6.40 per cent).
104
(3) Fat ix Dried Milk and Milk Po\yders.
Only a few analyses of these products were made, all in the laboratory of the Wis-
consin station. The results are presented in the following table:
Fat in dried milk and milk powders, determined by tuo methods.
Material.
Ether ex- '•
traction. !
Gottlieb.
Diflerence.
Milk flakes
Per cent.
0.31
.52
.35
17.04
29.13
7.75
25.37
Per cent.
0.80
1.74
2.19
17.32
31.66
8.41
27.43
Per cent.
+0 49
Nutrium
+ 1 ■^'^
Milcora
+1.84
-t- .28
Creamora
+*> 53
Tru-milk
+ .66
+2.06
11.49
12.79
+ 1.30
Incomplete as the investigation is. the results obtained show plainly that the ordi-
nary ether extraction of fat in dried milks give lower results than the Gottlieb method,
the differences ranging from about 0.5 per cent to 2.5 per cent in dried milks high in
fats, with an average difference of 1.30 per cent in favor of the Gottlieb method.
The accuracy of the results obtained by the Gottlieb method will be discussed under
the following heading.
(4) Results with the Gottlieb Method.
No reports of work with this method, aside from the results already presented with
condensed milk, were received by your referee. In our own laboratory a number of
samples of skim milk, buttermilk, and cheese were analyzed during the early part of
the year by G. von Ellbrecht and George A. Olson, comparative determinations being
made by the official asbestos extraction method, the Gottlieb method, and in some
cases by the Babcock centrifugal method. The results of these determinations are
given below.
Comparison of fat determinations made on different materials by three methods (Wisconsin).
Method.
Material.
Extrac-
tion.
Gottlieb.
Babcock.
Per cent.
0.091
.046
.155
Per cent.
0.143
.133
.192
Per cent.
Trace.
0.02
.02
Average
.097
.152
.02
.133
.153
.190
.403
.347
.379
.12
.11
.04
.159
.376
.09
New milk
3.85
3.81
4.61
3.95
3.82
4.74
Average
4.09
4.14
Cheese.. ..
36.10
37.20
35.11
36.50
37.32
36.75
Average ....
36.14
. 36.86
a A sample of buttermilk analyzed at the Cahfornia station gave the following results: Extraction
method, 0.85 per cent, GottUeb method, 0.90 per cent; Babcock test, 0.75 per cent.
105
In the analysis of cheese l)y the Gottlieb method 1 cc of hydrochloric acid and 9 cc
of water were added to al)oiit 2 grams of cheese and the mixture heated until a uniform
emulsion was obtained. This was then treated with 10 cc of alcohol and 25 cc each of
sulphuric and petrolic ether, shaking thoroughly after each addition, as in the case of
milk. The official gravimetric method for fat in cheese gave, on the average, 0.72
per cent lower results than the Gottlieb method modified to this extent that hydro-
chloric acid instead of ammonium hydroxid was used for bringing the cheese into an
emulsion, a modification that has been adopted by many European chemists.
The Gottlieb method is, as stated in the directions for cooperative work, especially
adapted for the analysis of milk low in fat, like skim milk and buttermilk. A large
amount of analytical work has been done with the method in Europe during the last
two years which shows conclusively that it gives correct results and that the fat
obtained by this method is pure butter fat. It follows, therefore, that all the fat is not
obtained in our official ether-extraction method or methods, apparently because the
fat is protected by a layer of sugar or nitrogenous substances, or both, which is formed
in drying the milk. On this point it is only necessary to refer to recent publications
on the subject of milk analysis, especially to T. S. Thomsen's « work, showing that
when milk proteids are peptonized prior to extraction there is no difficulty in reaching
as high results by extraction as by the Gottlieb method. (See also the following paper
by George A. Olson, on the examination of the fat obtained by the two methods). A
similar single experiment in our laboratory with a sample of condensed milk gave 9.66
per cent of fat by ether-extraction and 9.94 per cent by the official method after digest-
ing with one-tenth of a gram of pepsin per 250 cc of a 40 per cent solution for several
days (doubtless longer than was necessary) — i. e. , practically the same result as obtained
in the Gottlieb method. For the reasons stated, the referee would recommend that
the Gottlieb method be made provisional for the analysis of fat in milk.
The referee has been instructed by a unanimous vote of the Committee on Revision
of Methods to recommend that the conversion factor for protein in milk be changed to
6.38, which is the factor now quite generally uged by dairy chemists in this country
and abroad. All who are at all familiar with the literature on the subject know that
it is wrong to apply the factor 6.25 to dairy products, and this factor has been abandoned
by most chemists doing work in this line. It has been allowed to remain in our methods
only through conservatism, or through an oversight, because nobody has interested
himself in having it corrected. The referee would recommend, therefore, that the
factor be changed to that adopted by English chemists, 6.38; 6.37 or 6.40 would be
equally acceptable, were it not desirable to have methods and statements of analysis
always uniform in different countries, so far as possible, where this can be attained
without sacrificing accuracy of results.
Recommendations.
It is recommended that —
(1) The following method of extraction be adopted as provisional for the determina-
tion of fat in condensed milk:
Fat. — Extract the solid residue of about 5 grams of a 40 per cent solution of the con-
densed milk with ether in the usual manner, dry, leave tubes in the dish containing
about a liter of water for two or three hours, extract again for about five hours, and
determine fat as under milk. (Bulletin 46, page 54.)
(2) The Gottlieb method be made provisional, the following directions to be included
under determination of fat in milk:
Place about 10 grams of the milk in a 100 cc burette or eudiometer tube, add 1 cc of
concentrated ammonium hydroxid and mix thoroughly. Add the following chem-
icals in the order given: 10 cc of 92 per cent alcohol, 25 cc of washed ether, and 25
cc of petroleum ether [of a boiling point below 80° C, closing the cylinder with
a Mselkeritidende, 1905, 18: 359-365; Exper. Stat. Record., 1906, 17: 437.
106
a moist cork stopper and shaking the contents several times after each addition.
Leave for six hotirs or more; pipette off the clear fat solution into a small weighed flask
by means of a siphon drawn to a fine point, which is lowered into the fat sohition to
within 0.5 cm of the turbid bottom layer. Evaporate the ether solution in a hood, dry
in a steam oven for two to three hom's and weigh. In the case of new milk and dairy
products high in fat treat with a second amount of 10 cc each of sulphtiric and petrolic
ether to recover the last traces of fat.
(3) The conversion factor for protein in milk and dairy products be changed to 6.38
throughout the methods.
(4) The study of methods of analysis of condensed milk be continued, especially
with reference to the determination of lactose and sucrose in the sweetened product.
SUBEEPOET ON ANALYSIS OP DAIEY PEODUOTS.
By G. E. Patrick and M. Boyle.
With four of the five- samples of unsweetened condensed Inilk. reported in the follow-
ing table, one extraction by the Gottlieb-Roese method gave results a trifle higher than
did two extractions (with intermediate extraction with water) by the Adams paper coil a
method — and 0.2 to 0.25 per cent higher than by a single extraction by that method.
With one of the five samples — referee's sample Xo. 2930 — the result was quite
different; a single extraction by the Gottlieb-Roese gave 0.2 per cent less than was
obtained by even one extraction by the Adams. For this variation we have no expla-
nation at present. Forty per cent solutions of the samples were used for analysis.
Fat determined in unsweetened condensed milk by two methods.
Sample.
Adams method.
One ex-
traction,
20 to 30
hours.
Two extrac-
tions, with
intermediate
soaking in
water.
Gottlieb-
Roese
method,
one ex-
traction.
Xo. 2930 (referee's sample)
Per cent.
7.44
7.52
7.43
Per cent.
7.55
7.61
7.50
Per cent.
7.26
7.17
7.24
7.29
Average 7. 46 i 7. 55 7. 24
Second ether extract (grami , 0.0019-0.0028
No. 2528: |— ' —
Average
Second ether extract (gram).
No. 2530:
7.69
7.68
0. 0052-0. 0059
7.60
8.03
9.42
9.64
9.31
9.54
9.35 i
9.55
Average ' 9.36 , 9.58 9.62
Second ether extract (gram) r 0. 0065-0. 0079
No. 2610: =^
8 22
8.20
8.24
5.29
5.36
8.47
8.40
Average 8.22 I 8.35 8.43
Second ether extract (grami... , 0.0023-0.0049
a Analvst. 1865, 10: 46.
107
Fat determined in unsweetened condensed milk by two methods — -Continued.
Adams method.
Gottlieb'
Roese
method,
one ex-
traction.
Sample.
One ex-
traction,
20 to 30
hours.
Two extrac-
tions, with
intermediate
soaking in
water.
No. 2529:
Per cent.
6.94
6.95
6.87
Per cent.
7.00
7.06
6.99
Per cent.
0 7.12
b
17.17
6.92
7.02
0 0015-0. 0022
7.14
a In these two Gottlieb-Roese tests, 2 cc of water were added, and 2 cc more of alcohol than the direc-
tions require. It is possible that this change increased the result a trifle.
The referee's sample of sweetened condensed milk was lost by breakage in transit,
but the following determinations were made on other milks:
Fat determinations in sweetened condensed milk by two methods.
Adams method.
Gottlieb-
Roese
method,
one ex-
traction.
Sample.
One ex-
traction,
20 to 30
hours.
Two extrac-
tions, with
intermediate
soaking in
water.
No. 2611:
Per cent.
8.66
8.85
8.70
Per cent.
8.83
9.05
8.96
Per cent.
8.72
b
8.74
.
Average
8.74
8.95
0.0020-0.0031
8.73
Second ether extract (gram)
No. 2621:
a i
9.52
9.62
9.55
10.03
10.06
10.03
9.55
9.49
9.56
10.04
0.0050-0.0056
9.52
Second ether extract (gram)
These sweetened condensed milks were diluted with five times their weight of water,
before testing by the Adams method, so that the amount of the original sample on an
extraction paper was but little more than 1 gram. This high dilution magnifies errors
seriously, and one variable error, that due to ether extract from the paper, is unavoid-
able. We have never been able to obtain a uniform ether extract from blank papers,
the range being from 1 to 4 mg usually.^ In all of the work reported in this paper a
deduction of 0.0018 gram for this error was made from the first extraction and none from
the second. An error of 1 to 2 mg in the first or second extract, or in the two together,
is easily possible from this source; and on such small charges of the original material
even these small errors become of importance when expressed in percentage. Here,
therefore, in the error due to the extract from the papers, is a serious fault of the
Adams method, a fault which can be overlooked in ordinary milk analysis, but which
assumes importance when the method is applied to condensed milks, especially the
sweetened ones.
However, after making reasonable allowance for errors due to this cause, it appears
from the above work on the two samples of sweetened condensed milk that one extrac-
tion by the Gottlieb-Roese method gave about the same results as did one extraction
by the Adams method, and the query arises whether a second extraction by the Gott-
108
lieb-Roese "u-ould not have raised the results as much as did the second extraction by
the Adams. Unfortunately this comparison was not made.
In caiTying out the Adams method in this laboratory [Bureau of Chemistry] the milk,
properly diluted if need be. is spread upon an S. & S. ''fat-free" paper from a weighing
pipette, the paper is allowed to become sensibly dry at room temperatm-e, is then dried
for ten to fifteen minutes in the oven at 100°, immediately rolled in a coil, with two
very narrow strips of the same kind of paper between the layers (in place of the string
proposed by Allen o) , and extracted in a Knorr apparatus with water-free ether. This
apparatus probably acts less rapidly than does the Soxhlet upon such material, and for
this reason the time of extraction was extended to over twenty hours in the work here
reported.
Fat detenninatio72S ni huttermilk by two methods.
Adams method.
Gottlieb-
Rose
method,
one ex-
traction.
Sample.
One ex-
tra(?tion.
30 hours
. or more.
Two extrac-
tions, with
intermediate
soaMng in
water.
No. 2937:
Per cent.
0.681
.724
.781
Per cent.
0.730
.776
.830
Per cent.
0.843
b
.841
C-
Average
.729
.779
0.0020-0.0023
.842
Second ether extract (gram'>
This buttermilk is believed to have been produced by churning sour milk. It was
diluted with an equal weight of water before treating by the Adams method. The
Gottlieb-Roese method gave distinctly higher results than did the Adams, even with
two extractions, and the results on the same sample agree better. The Babcock
centrifugal method gave on this sample, in triplicate test. 0.80, 0.80. and 0.80.
One point observed in the working of the Gottlieb-Roese method seems worthy of
mention. When the line between the mixed ethers and the watery liquid is not clear
it can usually be made perfectly clear by adding a little sodium chlorid — 0.1 to 0.3
gram suffices. It must be added after the two ethers have been added and shaken
in, not before. Its effect is of course purely physical and perhaps any other salt
would do as well.
Instead of the original Gottlieb tube, we have used with much satisfaction Rohrig's
modification, & which is provided with a stopcock for drawing off the ethereal solution
of fat. But the stopcock should be lower down on the tube than Rohrig advises — i. e.,
at the 22 or 23 cc rather than the 25 cc mark.
a Analyst. 1886, 11: 72.
&A. Roln-iff, Zts. Xahr. Genussm.. 1905. 9: 531.
109
Lactose detenninations in unsweetened condensed milk.
Sample.
By copper
reduction,
Soxhlet
method.
By polari-
scope, clarifi-
cation by
mercury:
(acid
Hg(N03)2).
No. 2930, referee's sample;
Per cent.
10.01
10.02
10.06
10.06
Per cent.
y.
c . .
d
10.04
No. 2530:
a..
10.50
10.50
10.52
b
10.19
Average
10.51
No. 2528:
a. . .
10.73
10.65
10.68
1
b
\ 10. 57
Average
10.69
No. 2610:
a
10.14
10.15
10.16
b ,
i 9.97
Average
10. 15
No. 2529:
a
9.18
9.17
9.25
i 8.71
Average
9.20
No. 2531:
• 9.40
9.37
9.33
Ij
c
\ 9.00
Average
9.37
In the polariscope work, correction for the volume of the mercuric precipitate was
made according to Leffman and Beam — weight of proteids multiplied by 0.8 and
weight of fat by 1.075.
With five out of the six samples the results by the polariscope were lower than by
copper reduction; in three cases the deficit was serious, ranging from 0.32 to 0.49 per
cent. The results by copper reduction are believed to be the more trustworthy.
FAT DETEEMINATIONS IN OHEESE^ BY THE GOTTLIEB AND THE ETHEK-
EXTRAOTION METHODS.
By GeorcxE a. Olson.
It is conceded that higher percentages of fat are generally obtained by the Gottlieb
than by the present ether-extraction method. At the same time it has been questioned
whether or not this increased per cent of fat is not due to some substance or substances
other than fat, introduced by the chemicals used as solvents in the Gottliel:) method
or through their action.
A blank test of the chemicals used as solvents was first made and after evaporation
and drying no increase in weight was obtained. Comparative examinations of the fat
obtained by the Gottlieb and the ether-extraction methods were then made. Equal
110
quantities of a sample of old cheese were extracted by the two methods given, the
fatty solution was evaporated, and the fat dried for eighteen houi-s at 92° C. The
two preparations of fat had the following chemical and physical properties:
Determuiofion of fat in cheese by the Gottlieb and the ether-extraction methods.
CHEMICAL EXAMINATION.
Test.
Gottlieb
method.
Ether-ex-
traction
method.
Refractive index. 36= C 41 41. 0
38° C 40 40. 0
41° C 39 38.5
Specific gravitv. 69° C 0. 900 0. 8995
Kottstorfer number 2-30. 69 228. 34
Butyric acid per cent. . 5. 04 5. 04
Melting point a ° C. 29.5 29.5
PHYSICAL EXAMINATION.
Gottlieb method.
Ether-extraction method.
Clear; medium dark color; solidified at room tem- Cloudv; lishter color; not completelv solidified.
perature (22° C.i.
aWUey's method.
These results, which are the averages of several determinations, indicate such
slight differences in the chemical and physical properties of the fat obtained by the
two methods that it can not be definitely stated that the fats obtained by the two
methods are not identical. Wider differences may be obtained in the fat from the
same sample by either of the two methods of extraction. Since the two fats appear
to be similar in nature and true butter fat. it follows that quantitative results obtained
by the Gottlieb method sliow the true fat content of dairy products, and this is of
especial importance in the case of milk or other dairy products containing only small
amounts of fat.
DETERMINATION OE THE ACIDITY OE CHEESE.
By Alfred W. Bosworth. .
In the official methods under cheese analysis a provisional method for the determi-
nation of the acidity of cheese is given, which reads as follows:
To 10 grams finely divided cheese add water, at a temperatiue oi 40". until the
volume equals 105 cc: agitate vigorously and filter. Titrate portions of 25 ec of
filtrate con-esponding to 2.5 grams of cheese with standardized solution of sodium
hychoxid. preferably one-tenth normal. Use phenolphthalein as indicator. Express
amount of acid as lactic.
The following modification is proposed:
Extract 25 grams of finely divided cheese with 200 cc of water at 55° C. decant the
supernatant liquid onto a cotton filter, and complete the extraction with sticcessive
portions of water until nearly 1 liter is collected. Make up to 1 liter, shake, and
titrate 100 cc with twentieth-normal sodium hydroxid. using phenolphthalein as an
indicator. The figine obtained is to be multiplied by 20. which will give the acidity
of 100 gi'ams of cheese expressed as tenth-normal alkali.
The lactic acid present in cheese was determined by grinding 5 grams of cheese
with sand, extracting with watf-r acidulated with sidphmic acid, and extracting the
resulting solution with ether. A lactic acid determination in this ether extract gave
Ill
0.405 per cent. At the same time a water extract was prepared, as explained above,
and a determination of lactic acid in it gave 0.410 per cent. This acid was in com-
bination, and therefore the test for free lactic acid gave no results. As free lactic
acid is never found in cheese, there seems to be no reason why the acidity should be
expressed as lactic acid, as the official methods direct.
The lactic acid is present as calcium lactate. The bacteria present in the cheese
produce lactic acid from the milk sugar. This lactic acid as developed splits off cal-
cium from the calcium paracasein and the insoluble calcium phosphate, forming calcium
lactate, soluble calcium phosphate, and free paracasein. The calcium lactate is
neutral, and the free paracasein, being insoluble, does not affect the acidity as deter-
mined. It is the soluble calcium phosphate which causes the acidity.
Tables 1 and 2 show how the acidity of the cheese increases with the increase in
solubility of the phosphoric acid. The acidity as given in these tables is the maxi-
mum acidity of the cheese examined, and it will be noticed that this point coincides
with the time when the water-soluble phosphoric acid becomes 100 per cent of the
total inorganic phosphoric acid in the cheese. An increase in the soluble calcium is
noticed after this, but it does not seem to affect the acidity.
Table 1. — Acidity determinations in Camembert cheese.
Age of cheese.
Acidity of
100 grams
of cheese.
Water-
soluble
phosphoric
acid.
Water-
soluble
calcium
oxid.
cc tenth-
normal
alkali.
24.0
60.0
100.0
Per cent total
inorganic
P2O5.
73.53
94.87
100.00
Per cent total
CaO.
39.02
82.22
16 hours
97.62
Table 2. — Acidity determinations in Cheddar cheese.
Age of cheese.
.\cidity of
100 grams
of cheese.
Water-
soluble
phosphoric
acid.
Water-
soluble
calcium
oxid.
When whev was drawn. .
cc tenth-
normal
alkali.
48.0
116.0
130.0
190.0
Per cent total
inorganic
P2O5.
50.98
69.07
73.59
100.00
Per cent total
CaO.
27.93
6 hours
57 33
62.82
2 weeks . .
80. 16
Mr. Patrick. This paper suggests the question, Should we deter-
mine only the water-soluble acidity of cheese ? O. Jensen has recently
published an article in which he advises the determination of the
acidity of the entire cheese, casein included. He recommends rub-
bing up the cheese with water and titrating the entire mass. There
is certainly a difference of opinion as to which is the better method.
Mr. BoswoRTH. Such a determination would not be uniform, for
the reason that the amount of free casein in two different cheeses
might vary considerably, and the acidity of the water-soluble por-
tion, depending upon conditions which are not under control in the
manufacture of cheese, would not be determined.
112
^Ir. CocHRAX. I noticed in the report on condensed milk that the
sucrose determination is to be investigated. This subject has
interested me of late, and I have been using acid mercimc nitrate
to invert the sucrose in condensed milk with satisfactory results.
This method, known as the Harrison method, was published in the
Analyst about a year ago, but I have seen no reports of work done
with it.
^Ir. Patrick. I have found in using acid merciu-ic nitrate for
sweetened condensed milk that there is danger of inversion taking
place to some extent before the fu^st polarization is made unless tliis
is done immediately after filtration. A neutral solution of mercuric
nitrate, proposed by Patein and Dufau, appears to be better than
the acid solution.
EEPOET ON POODS ATO lEEDING STUFPS.
By J. K. Haywood. Referee.
During the j^ast ten years the attention of the referee on foods and feeding stuffs has
been directed toward the collection and comparison of methods for determining
moisture, ash, ether extract, crtide fiber, albumenoids, starch, pentosans, and galactan.
The official methods as they now stand represent the best of the above methods, from
which the poorer methods have been gradually excluded. We all laiow that most of
these methods give only approximate results, and that even these can not be obtained
unless the most minute details of the method are followed to the letter. However,
they are the best obtainable in the present state of our knowledge, and the details are
as exact as a long series of comparative studies can make them. It seems, therefore,
that a further comparison of these methods by different members of the association
would be almost useless until decidedly new facts are ascertained regarding the separa-
tion and determination of cattle-food constituents.
When the last report was made on this subject, in 1903, the referee brought out
certain facts in regard to crude fiber, which seem, however, to merit further study.
Besides this, a method of determining a new constittient of cattle foods was recently
published, i. e.. methyl pentosans, which gives promise of being of some value. It
is along these two lines, then, that the referee has worked. Samples for compara-
tive study were not sent out. as it was thought best to devote all the time that the
referee could spare to formulating the methods, so that they could be comparatively
studied next year.
In 1903 a comparatiA'e study was made of the present official method of determining
crude fiber; of the Ivonig method, by boiling with glycerol sulphuric acid for one
hour, and of the modified Konig method, by boiling with glycerol sulphuric acid
and then with 1.25 per cent sodium hydroxid. It was found that the present official
method gave a fiber nearly free from albuminoids, but containing a large amount
of pentosans; that the Konig method gave a fiber nearly free from pentosans, but
containing considerable albuminoids, while the modified Konig method gave a fiber
containing only negligible quantities of both pentosans and albuminoids. It was
suggested by the referee, however, that the low results on fiber obtained by the modi-
fied Konig method were not only due to getting rid of all pentosans and albuminoids,
but that a hydroH-tic action on the cellulose of the crude fiber was exerted by the acid
in the presence of glycerol, at a temperature of 131°-133° C. This suggestion was
tested upon pure cotton cellulose, and when treated by the Konig method there was a
loss of 12.35 per cent, due evidently to hydrolytic action. The referee was not willing
113
to condemn the Konig method on these figures alone and. recommended a further
study of the subject by the succeeding referee.
In working upon this subject as pure cellulose as possible was prepared by heating
absorbent cotton, first with dilute sulphuric acid, then with dilute alkali, washing,
and drying to constant weight. This cotton was used in all subsequent determina-
tions. To ascertain just how much loss in weight there is by our present official
method, a 1-gram portion of the cotton was boiled one-half hour with 1.25 per cent
acid, under which treatment it lost 2.70 per cent in weight. Another 1-gram portion
was then boiled one-half hour with 1.25 per cent sodium hydroxid, by which treat-
ment it lost 17.06 per cent in weight. It is therefore evident that by our present
official method there is a loss in pure cellulose of 19.76 per cent. With glycerol sul-
phuric acid and one hour's boiling there was a loss of 26.31 per cent, while with the
modified Konig method there was a loss of 43.37 per cent. It is evident from the above
that the modified Konig method is not an improvement on the old official method, since
the great loss from hydrolysis of the cellulose makes an error in the minus direction
greater than the plus error caused»by the presence of pentosans in the crude fiber pre-
pared according to the present official method.
It was hoped that a satisfactory method for determining crude fiber might be devised
by using the Konig method and boiling only one-half hour instead of one hour with
the glycerol sulphuric acid. This was done and it was found that the pure cellulose
lost 29.40 per cent in weight. This loss in weight is almost 10 per cent greater than
by the present official method. Such a showing would hardly entitle the method to
consideration in comparison with the present official method, and it is therefore rec-
ommended that further work on the Konig method for determining crude fiber be
abandoned.
The next subject studied by your referee was the determination of methyl pentosans
in the presence of pentosans, by a method recently published by Ellett and Tollens. «
This method is based on the following jDrinciples: First, it was shown by Ellett and
Tollens that in certain vegetable materials both pentosans and methyl pentosans are
present. On distillation with hydrochloric acid both of these distil over, the first as
furfural, and the second as methyl-furfural. On precipitation with phloroglucol,
the furfural-phloroglucid and the methyl-furfural-phloroglucid are precipitated. In
the past both of these have been weighed together and calculated as so much
phloroglucid, thus causing an error in the pentosan determination. Ellett and Tollens
have worked out a method for the separation of these compounds, based on the solu-
bility of the methyl-furfural-phloroglucid in alcohol, and the insolubility of the
furfural-phloroglucid in the same medium. It is claimed by these authors that the
solubility of furfural-phloroglucid in alcohol is practically negligible, so that the amount
of material extracted from furfural-phloroglucid and methyl-furfural-phloroglucid
represents the methyl-furfural-phloroglucid alone. It was with an idea of testing
this method for determining methyl-pentosans and the underlying principles of the
same that the following experiments were^made.
It was first necessary to determine whether Ellett and Tollens were correct in saying
that the furfural phloroglucid was practically insoluble in alchohol. For this purpose
weighed quantities of arabanose (in quadruplicate), were distilled with hydrochloric
acid, precipitated with phloroglucol, dried four hours, and weighed in a weighing
bottle according to the official method. The weights of phloroglucid obtained, using
two different weights of arabanose, in quadruplicate, were as follows:
0.0809, 0.0771, 0.0777, 0.0790 gram
0.2313, 0.2365, 0.2392, 0.2366 gram.
« J. Landw., 1905, 53 [1]: 13.
31104— Xo. 105—07 8
114
These were extracted with alcohol in the following manner: The weighed Gooch
crucibles containing the dried phloroglucid were placed in a 100 cc beaker, and 30 cc
of 95 per cent alcohol at 60° were poured into the Gooch. These beakers were then
placed in a water bath and extracted for ten minutes at 60° C. All alcohol was sucked
fi-om the Gooch by a suction pump, poured back into the beaker, and passed through
the filter again to catch paiticles of asbestos fiber. This alternate extraction and filtra-
tion was repeated three times. The last extract was only slightly colored.
The Gooch crucibles were finally di-ied in the water oven for two hours and again
weighed in weighing bottles. The difference between the first and second weighings
of the phloroglucid precipitates, representing the amount extracted by alcohol, were
respectively as follows:
0.0039. 0.002S. 0.0033. 0.0032 gram
^^ 0.0033. 0.00i5. 0.0034. O.OOiO gram. .
This is an average of 0.0036 gram, a much greater amount than that which EUett
and Tollens fotind to be extracted, namely (0.0012 gi-am i. so that it would appear that
a correction of the methyl -farfural-phloroglucid obtatned,by subtracting this amotint
of material, i. e.. 0.0036 gram, would be necessary, especially where the methyl-
pent osans are very low. as they usually are with respect to the pentosans present.
Since it might occasionally be necessary to extract the phloroglucid precipitate more
than three times to dissolve all of the methyl-furfural-phloroglucid. an experiment was
next undertaken to ascertain what amount of furfural-phloroglucid is extracted by
treating the precipitate five times with alcohol, as just described. For this purpose
quadruplicate samples of arabanose were weighed out and determined as above; the
amount of furfural-phlnroglucid obtained was as follows:
0.0S23. 0.0835. 0.0835. 0.0800 gram.
And the amount of the same extracted by five treatments with alcohol was:
0.0043. 0.00:38. b.0041. 0.0040 gram,
or an average of 0.0041 gram. It will thus be seen that the difference between the
amounts obtained by three and five extractions is so small (0.0005 grami that it is
immaterial which number is made.
There is a very important point which is emphasized by this ver^" slight difference in
the amount extracted fi'om the phloroglucid precipitate by three and five treatments.
If the amount extracted from this phloroglucid precipitate by alcohol really represents
the solubility of furfural-phloroglucid in alcohol, we would expect nearly twice as much
to be extracted by fi^'e treatments as by three, btit such is not the case. Further, on
extraction with successive ponions of alcohol the first extracts are much darker colored
than the later ones, showing that more material is extracted in the beginning than sub-
sequently. This again appears to indicate that the amount of material extracted does
not really represent the solubilitj^ of the furfural-phloroglucid in alcohol. For these
two reasons the writer is inclined to believe that either a verj- small amount of some
secondaiy product extractable by alcohol is formed dm-ing the distillation of a
pure pentose, in this case arabanose, or that some of the phloroglucol used for the pre-
cipitation is occltided by the precipitate, is not entirely washed out. and is finally
extracted by the alcohol. For various reasons, diflicult to explain, but evident to one
who has used the method a few times, the referee inclines to the latter view. This is
a matter which merits further study. However, whether this alcohol extract repre-
sents the solubility of furifui-al-phloroglucid in alcohol, is a secondary- product, or is
occluded phloroglucol, is not of special significance in separating methyl-furfural-phlo-
roglucid from furfural-phloroglucid. The fact remains that a certain amount of some
material is in the fm-fural-phloroglucid which is extracted by alcohol and is fairly con-
stant in amount. This amount must be subtracted from the methyl-furftiral-phloro-
glucid found before even fairly correct results are obtainable.
It was thought possible that on drying the phloroglucid for two hom-s extra, after
extracting with alcohol, some oxidation of the precipitate would take place, so that
115
the difference in weight of the phloroghieid ))ef()re and after extraction, according to
the above method, would not give the true weight of the material extracted by the
alcohol. To test this point, four samples of phloroglucid," prepared by distilling
arabanose as described above and precipitating it with phloroglucol, were treated just
as the precipitates were treated in the previous case, except that alcohol was not placed
on them, i. e., they were first dried for four hours, weighed in a weighing bottle, then
removed from the drying oven and placed on filter pumps, where air was sucked
through them for about the same period as when alcohol was used for the extraction.
The phloroglucid precipitates were then returned to the water oven and dried for four
hours instead of two, so that if any oxidation took place it would be apparent.
The precipitates were finally removed from the oven and again weighed in weigh-
ing bottles. Working in this way one lost 0.0005 gram, two weighed exactly the same,
and one lost 0.0010 gram. It is therefore evident that the phloroglucid precipitates
do not oxidize and so gain in weight by the extra heating after the extraction with
alcohol. It appears from this work that it would be safer to continue the second dry-
ing for four instead of two hours, since in two or three cases where the samples were
dried only an extra two hours after having been treated as above, they had gained
about a milligram in weight, but on further heating they returned to the original
weight, showing that it was not an oxidation process that caused the gain.
It might be easier to evaporate the alcoholic extracts and weigh them directly,
rather than to get the weight of the material extracted with alcohol by difference.
This was done in all the above cases, and the alcoholic extracts by difference gave the
following weights: 0.0039, 0.0028, 0.0033, 0.0032, 0.0033, 0.0045, 0.0034, 0.0040, 0.0043,
0.0038, 0.0041, 0.0040. By direct weighing, the following results were obtained:
0.0075, 0.0054, 0.0055, 0.0052, 0.0082, 0.0084, 0.0085, 0.0087, 0.0087, 0.0075, 0.0076,
0.0073.
From these determinations it appears that the average weight of the extract obtained
from furfural-phloroglucid by indirect weighing was 0.0037 gram, while the average
weight obtained by direct weighing was 0.0074 gram, just twice as much. This result
was entirely unexpected, and up to the present time the writer is unable to explain it.
At first it was thought that the 0.0074 gram was the true weight of the material extracted
from the furfural-phloroglucid, and that the weight 0.0037 was lower than it really
should be because of the oxidation of the phloroglucid in drying during the extra two
hours. This supposition, however, was proved to be wrong by the experiment quoted
above, showing that the phloroglucid does not gain in weight on further drying. It
now seems that the only possible explanation of this amount of material in the residue
from the alcohol extraction is that some reaction takes place between the material
extracted and the hot alcohol, resulting in a heavier compound. This is also a ques-
tion which must be studied before the method can be considered as well established.
Since Ellett and Tollens have made so many determinations showing that the
methyl-furfural-phloroglucid is soluble in alcohol, it was not deemed necessary to test
this point. In view of all the facts as stated the referee would recommend that the
following slightly modified method of Ellett and Tollens be tested next year by the
association:
Method for Determining Pentosans and Methyl Pentosans.
Proceed as in the determination of pentosans by the official method until the phloro-
glucid precipitate has been dried for four hours and weighed. Place the Gooch cru-
cible containing this precipitate in a 100 cc beaker and pour into the Gooch 30 cc of 95
per cent alcohol heated to 60°. Place the beaker for ten minutes in a water bath
o It should here be mentioned that the phloroglucol used was an imported article,
said to be free from diresorcol. It was further purified by the official method of the
Association of Official Agricultural Chemists, Bureau of Chemistry Circular 30, p. 5.
116
heated to 60°. Remove the beaker and Gooch and suck from the Gooch all alcohol
remaining therein with a suction pump. Repeat this alternate extraction and sucking
dry of the precipitate three to five times, according to the color of the filtrate obtained.
After the final extraction place the Gooch crucible in a wateroven and dry foiu- hours,
making the final weighing in a closely stoppered glass weighing bottle as described in
the official method for pentosans.
The difference in weight between the furtural-phloroglucid plus methyl-furfural
phloroglucid first obtained and the furfural-phloroglucid remaining after extraction
with alcohol, minus 0.0037, represents the amount of methyl-furfural phloroglucid
pTesent, from which the original pentose (calculated as rhamnose) can be calctilated by
the following formula :
Rhamnose=(Phi (165. — (Ph)^ (1.84)+0.010.
Ph equals the weight of methyl-f urfural-phloroglucid : Rhamnosan equals rhamnose
mtiltiplied by 0.8. A table will be found on page 19 of the article of EUett and Tollens,«
in which attention is directed to an error in the calculation of 0.028 and 0.029 gram of
the methyl-f urfm-al-phloroglucid to rhamnose.
To obtain the weight of pentosans, subtract the final weight of methyl-fiuiiual-
phoroglucid obtained above from the weight of the mixture of methyl-f urfural-phloro-
glucid and fiu-fiu-al-phloroglucid and calculate accordingly to Ivi-ober's tables, or
according to the formulas given in the present official methods for pentosans.
Working in this way the following results were obtained in quadruplicate on samples
of gum arable and tragacanth:
Quadruplicate determiiiations of pentosans and methyl pentosans made by the proposed
method.
Detenu inations.
Pentosans.
Gum
a r able.
Traga-
canth.
Per cent. , Per cent.
26.18 ; 37.0.3
25.00 36.94
25. 88 38. 00
25. 36 36. 92
Determinations.
Gum Traga-
arabic. canth.
Methyl-pentosans (as rham-
nosan)
Per cent. Per cent.
3.28 1 4.38
3.54 I 4.54
3.54 ; 4.54
3.58 ' 3.98
EEPORT ON SUGAE,
By C. A. Browxe. jr.. Referee, and J. E. Halligax. Associate Referee.
The work of the referee and associate referee upon sugar during the past year has been
substantially along the lines recommended by the association at its last meeting and
has comprised (1) work upon the more special methods of anah sis in their relationship
to sugar chemistry-; (2^^ work upon ptu-ely chemical methods: and 1 3 ) work by a number
of collaborators upon methods for the analysis of cane molasses, massecuites. and
sugars.
In the investigation of special methods the work has been confined very largely to a
study of the organic constituents of cane molasses. Methods for the estimation of
nitrogen in the numerous forms under which it occui's in molasses have been com-
pared, also methods for the determination of the nonfermentible sugars — a matter of
considerable importance to distillers. The application of a method for analyzing
sugar mixtures, read at the last meeting, has been extended to a wide range of sugars
and carbohydrate bodies with a fair degree of success.
The work upon chemical methods has been continued l^y Mr. Munson. with the
cooperation of ^Ir. Walker, the result being that we have a table for the estimation of
a Loco cit.
117
dextrose and invert sugar, both alone and in the presence of sucrose, under perfectly
uniform conditions of analysis. Mr. Walker has lately added to this table columns for
the estimation of lactose and maltose, and it is hoped thatthe work may be extended
to other commonly occurring reducing sugars.
The present report will be limited entirely to a discussion of the coojDerative work Ijy
various chemists upon a low grade massecuite, molasses, and sugar, the work of the
referee on special analytical methods and of the associate referee on molasses methods
having been combined. The three samples sent out for analysis were obtained from
the Louisiana Sugar Experiment Station of New Orleans and represented the final
products from sugar cane grown at this station the previous year. In the circular
letter sent out with these samples the following instructions were given:
(1) WATER AND TOTAL SOLIDS.
Test as many of the following methods as possible:
(a) The official methods, page 27, Bulletin No. 46, Bureau of Chemistry.
(b) Drying 2 grams ten hours at temperature of boiling water, the loss at the end of
this time being taken as water, without regard to constancy in weight.
(c) Drying in vacuum at 70° constant weight. The massecuite and molasses should
be dissolved in an equal weight of water and about 2 grains of the solution weighed out
upon sand or asbestos.
(2) REDUCING SUGARS.
(a) Determine reducing sugars as dextrose according to the method of Allihn, Bulle-
tin 46, page 35.
(1) Upon the filtered solutions without clarification.
(2) Upon the filtered solutions after clarifying with 1 to 2 cc of lead subacetate.
(3) Upon the filtered solution after clarifying with an excess of lead subacetate, this
excess to be removed by sodium carbonate (official method, Bui. 46, p. 33), or by
potassium oxalate (Sawyer, recommendation 4, Cir. 26, p. 5).
(6) If desired compare the method of Allihn with that of Soxhlet or any other method
preferred by the cooperator.
(3) SUCROSE.
Determine sucrose by the following methods:
(a) The official method by inversion (Clerget, p. 39, Bui. 46). The normal weight
is made to 100 cc without dilution.
(6) By dilution (using Clerget's method as before).
(1) According to Sawyer, Circular 26, Bureau of Chemistry, page 5. Note carefully
recommendations 1, 2, 3, and 5.
(2) According to Geerligs. A portion of well-mixed sample is dissolved in 1 to 4
parts by weight of water without heating. A normal weight of this solution is clarified,
made up to 100 cc, and polarized in the usual way. If the solution is too dark to read
in the 200 mm tube, read in the 100 mm tube.
(c) Compare the optical methods for sucrose with the official gravimetric method,
Bulletin 46, page 39.
In reporting the results the cooperators are requested to report all analytical data as
fully as possible — the methods used, dilutions employed, temperatures of polarization,
manner of determining reduced copper, etc., in order to afford every facility for com-
paring the results.
It is also urged that the work upon the samples be begun, if possible, immediately
upon their receipt to avoid the liability of changes in composition through fermentation.
C. A. Browne, Jr.,
Referee on Sugar (Special Analytical Methods).
J. E. Halligan,
Associate Referee on Sugar (Molasses Methods).
Ten chemists signified their willingness to cooperate, and reports in whole or in part
were received from five.
Determination op Total Solids.
The results o])tained by four of the chemists upon total solids are given in Table 1.
118
Table 1. — Determinations of total .solids in massecidte, sugar, and molasses.
Method ;
.1
H. P. Agee. J. A. Hall, jr.,
sugar experiT sugar experi-
ment station. ment station.
New Orleans, La. New Orleans, La.
J. E. Halligan.
agricultural ex-
periment station.
Baton Rouge. La.
S. F. Sherwood.
Bureau of Chem-
istry. Washing-
ton. D.C.
i
U 1
i- X
i^- is
1 li 1^ 1
£ i"^ i w £
t'
0
2 grams. 9S°
Hs.
10
f 10
12
16
p.ct. p.ct.
82.09 95.65
82.40 95.63
8L 80. 95. 28
80.98 9.=S.09
p.ct. P.ct. P.ct. P.ct.
78.66 82.12 95.87: 79.16
78.5a 81.40 1 77.27
77.69 8L30 76.86
76. 66 Increase after 12
P.ct. P.ct.
82.20 95.42
P.ct.
77.66
78.76
78.04
77.56
P.ct.
82.21
81.59
81.30
80.54
79.88
79.66
79.30
81.85
81.37
80.70
79.54
25°
85.08
84.88
p.ct.
95.40
P.ct.
77.93
77.86
77.41
76.66
2 grams on sand, 98°. . ^
20 80.58 94.85
25 80.20
hours
76-06
75.89
75.65 ...
....
75.68
[30
f 10
12
16
[so
Vs.
3
5
8
13
20
. 40
80. 10
76.10
75. ''O
i
.1 82.22
......
78.82 84.20
77.88 83^=54
80.02
78.74
78.18
75.54
78.68
2 grams with water
81 Rfi
78.02
on pnmice stone. <
98°. T V .....
81.43
77.47
82.80
81.08
77.06
45° '45°
85.73 98.76
85.56 98.65
45°
81.06
80.96
65°
79.51
79.28
25° ' 25°
96.98 81.98
96.79 81.34
Hempel desiccator in
84 59 96 77 81 3"^
vacuo, 2 grams dis-
84.27 96.69 81 04
solved on paper or
asbestos
i
40°
82.82
82.26
40°
79.26
78..^
i
1
Samples taken from Hempel desiccator and dried 8 hours in vacuo, 6S°
74 95. 9S 79.20
Mr. E. B. Holland, of the Massacliusetts agiicultiu'al experiment station, obtained
by a method of drying on quartz sand for twenty hours at 75° and then three hours
at 100°, 82 per cent of solids in the massecuite, 95.27 per cent in the sugar, and 77.50
per cent in the molasses, these results comparing well with the 2-gram ten-hour method.
Some very interesting sets of experiments were submitted by Mr. B. L. Hartwell.
of the Bhode Island experiment station, showing the effects of dr^'ing in a vacuum
oven under varying conditions of temperature and pressure. During the diying.
air which had passed thi"ough concentrated sulphtu'ic acid a number of times was
drawn through the apparatus. Jja. one set of these experiments the dr%'ing was con-
ducted for successive periods with phosphorus pentoxid placed in a dish inside the
bath. The results of the experiment are given in Table 2.
Table 2. — Determinatimi of solids in massecuite and molasses by drying for consecutive
periods under low pressure over phosphorus pentoxid uith an air current drawn thorugh
five sulphm^ic and u-ash bottles.
[B. L. Hartwell and P. H. Wessels. .Agricultural Experiment Station. Kingston. R. "^.]
Period.
Temperature.
Pres-
sure.
Time.
Massecuite.
Molasses.
I.
II.
I. IL
1
2
3
4
63°....
58°-...
62°-...
70°--..
88°
mm.
16
82
95
85
48
Hours.
17
18
20
16
16
Per cent.
82.95
82.53
82.27
82.03
80.97
Per cent.
82.86
82.52
82.19
81.99
80.89
Per cent.
79.06
78.64
78.43
78.23
77.17
Per cent.
79.28
78.94
78.70
78.45
77.42
In commenting upon these results Mr. Hartwell says: '"The effect of dr^dng at 88°
illustrates hoAV readily decomposition takes place when the temperatm'e exceeds 70°
and it is perhaps a question whether the gradual loss which has taken place by the
continued drying is not due to some extent to decomposition rather than to moisture."
119
The results obtained by Mr. Hartwell al'tc^r drying fifty-fivo hours at 62° phow very
close agreement with those olitained l>y Mr. Halligan after forty days drying in the
Hempel desiccator. These results also correspond closely with those obtained by
the different chemists by drying 2 grams for ten hours at 98°. The purpose of the
work this year was to ascertain the most satisfactory method for determining moisture
and solids in sugar products under ordinary laboratory conditions, and nearly all the
cooperating chemists seem agreed that the 2-gram ten-hour method fulfils the con-
ditions most perfectly. The results obtained this year, as well as during previous
years, by this conventional method and the more accurate process of drying under
diminished pressure at 60°-70° show a fairly close agreement. The method seems
well adapted for routine laboratory work, and it is recommended that it be adopted
provisionally by the association. Where vacuum ovens are available and greater
scientific accuracy is desired, the method of drying upon some absorbent material
at a lower temperature under diminished pressure is of course to be preferred.
The methods of drying on sand and pumice stone at the temperature of boiling water
gave continual losses in weight on prolonged drying, though one chemist reported an
increase in weight after twelve hours heating. Drying solutions of the materials upon
paper or asbestos in a Hempel desiccator at the ordinary laboratory temperature will
remove practically all of the moisture after standing many weeks. The process, how-
ever, is slow and not adapted for general work. Drying in a Hempel at 60° to 70°
hastens matters considerably, the removal of moisture after one day at this tempera-
ture exceeding that obtained on thirteen days standing in the cold.
A difficulty is occasionally experienced by chemists in drying sugar-containing
materials which are acid, owing to the rapid inversion and decomposition which set in.
Pellet a advises the addition of a drop or two of ammonia in all cases before drying, to
neutralize any such acidity, and this precaution seems to be worthy of attention.
Determination op Reducing Sugars.
The results upon reducing sugars as determined gravimetrically by Allihn's, and
volumetrically by Soxhlet's, method are given in Table 3. For ease of comparison
all determinations are reported in terms of dextrose, though the expression of results
as invert sugar would be more accurate.
Table 3. — Determinations of reducing sugars in massecuite, sugar, and molasses.
Analyst.
Reducing sugars as dextrose, Allihn's method.*
No clarification,
solution filtered.
Mas-
se-
cuite.
Su-
gar.
Mo-
Clarified with
1-2 cc of lead sub-
acetate.
Mas-
se-
cuite.
Su-
gar.
Mo-
Clarified with an
excess of lead
subacetate.
Mas-
se-
cuite.
Su-
gar.
Mo-
Reducing sugars
as dextrose, the
Soxhlet volume-
tric method.
cuite.
Su-
gar.
Mo-
las-
H. P. Agee, sugar experi-
ment station, New Or-
leans, La
J. A. Hall, jr., sugar experi-
ment station, New Or-
leans, La
J. E. Halligan, agricultural
experiment station. Ba-
ton Rouge, La
S. F. Sherwood, Bureau of
Chemistry, Washington,
D.C
P.ct.
26.60
27.24
26. 91
t23. 70
P.ct.
7.22
7.22
7.74
7.14
P.ct.
3L74
31.42
32.50
31.54
P.ct.
25.62
26.28
26.84
622. 89
P.ct.
7.18
6.57
7.49
6.82
P.ct.
31.00
30.98
32.18
30.84
P.ct
25.04
23.00
24.
P.ct.
7.04
6.92
7.13
P.ct.
P.ct.
P.ct
28.54: 26.66
27.90
29.61
26.75
P.ct.
7.10 32.68
I
6.94 33.11
6.79' 32.20
* Reduced copper calculated from weight of cuprous oxid by all chemists,
t Low results due to leakage of molasses from sample during shipment.
« International Sugar Journal, 1906, p. 509.
120
A comparison of results shows that with the increasing addition of lead subacetate
for clarification, a very marked falling off in the copper reducing power takes place in
all the products tested. This, of course, brings up again the old question: '•Should
solutions be clarified or not before determining reducing sugars; and if so, how should
they be clarified?" The question also arises: "Do these copper-reducing bodies,
precipitated by lead subacetate, not belong to the non-sugars, and should they not be
removed? " Pellet in numerous articles has contended that these precipitated reduc-
ing bodies consist largely of levulose, and the referee is also inclined to the belief that
considerable quantities of reducing sugars are precipitated by the lead subacetate, not
simply fi'om the changes in reducing power and polarization — since this might result
from the removal of reducing, optically active bodies which are non-sugars — but also
from the decreasing yield of alcohol which fermented molasses solutions give after hav-
ing been clarified with subacetate. There is no doubt that we have here one of the
principal sources of the discrepancies found in determinations of reducing sugars, and
for the sake of simplifying conditions the discontinuance of the lead subacetate solu-
tion entirely in clarifying solutions for the determination of reducing sugars would
seem in many ways desirable, using for the purpose of removing slime and suspended
impurities some inactive agent such as alumina cream. The point is one which
should be fully investigated by next year's referee since the questions involved are of
the greatest importance in commercial work. In this connection it may be noted that
the English and French sugar chemists have not been uniformly favorable to the lead-
subacetate clarification in the determination of reducing sugars, while the German
chemists have been usually in favor of this preliminary treatment.
One very serious disadvantage of not clarifying our sugar solutions will be the greater
care necessitated in the determination of the reduced copper, owing to the gi-eater
liability of the precipitated cuprous oxid being contaminated with organic and mineral
matter. This is a frequent source of error even after clarification, nearly every chemist
having had the experience at times of the cuprous oxid coming down green or yellowish
colored, and retaining gelatinous organic or mineral impurities which render filtra-
tion difficult. Even when the precipitate is of the proper color there is no certainty
of its being free from contamination, especially in case of saccharine products of low
pm*ity. To illustrate this point more fully, there are given in Table 4 a few results
taken from a large number of comparative analyses made by Mr. S. F. Sherwood and
Mr. M. H. Wiley at the Bureau of Chemistry upon a variety of products.
Table 4.
-Comparison of methods for estimating reduced copper.
Material.
Reduced copper.
Analyst.
From
weight of
From
weight of
Volu-
metric
Remarks.
cuprous
cupric
method
oxid.
oxid.
(Low).
Gram.
Gram.
Gram.
fMolasses resid-
0.3753
0. 3594
0. 3494
Precipitate difficult to
uum.
filter.
1
....do
.3905
.3634
.3470
Do.
S. F. Sherwood, Bureau of
...-do
.2517
.2348
.2242
Do.
Chemistry* Washington,
....do
.3287
.3130
.3034
Do.
D. C. '
...-do
.3291
.3134
.3029
Do.
....do
.276^
.2698
.2688
do
.2709
.4619
.2620
.2612
.4617
do '.-
.2449
.2444
-...do
.1251
.1257
Beer
.0755
.0746
.4628
.07.53
.0748
.4520
M. H. Wiley, Bureau of
.. do
Chemistry, Washington,
D. C.
Com juice
.3360
.3134
Do.
Malt extract
.3322
.3048
Large amount of pep-
tones in extract.
do
.3160
.2093
.2933
.1934
Do.
i....do
Do.
121
In the results obtained upon the molasses residuum the precipitated cuprous oxid,
after weighing, was ignited in a muffle and weighed as cupric oxid according to the
method of the Fr(>nch chemists; the cupric oxid was then dissolved in nitric acid and
the copper estimated l)y the volumetric method of Low. The results show a contam-
ination of the cuprous oxid with organic matter as shown l)y the differences in copper
as calculated from the sul)oxid and oxid and with mineral matter as shown by the
dilferences in copper as calculated from the oxid and by the volumetric method.
With solutions of pure sugar and such liquids as beer, where the organic matter
consisted largely of carbohydrates, the calculation of copper from the weight of cuprous
oxid gave accurate results. In the case of the malt extracts, which contained added
peptones, the precipitated cuprous oxid seemed to act somewhat as the copper hydrate
of Stutzer's reagent and to carry down a considerable amount of albuminoid matter
from solution; in the case of the molasses the precipitated copper seemed to be in
partial combination with certain nitrogenous bases such as xanthin. The investiga-
tion of a suitable clarifying agent which will remove such contaminating bodies with-
out precipitating the sugars, and the study of a rapid accurate method for estimating
reduced copper, such perhaps as the electrolytic method with rotating anode, are
offered as recommendations for next year's work.
The volumetric method of Low, suggested by Mr. Munson as a provisional method
in his report of last year, has been tested by a number of the chemists with uniformly
favorable results. Mr. E. B, Holland writes: "The method as a whole after a thorough
trial proved perfectly feasible, being fairly simple and of easy manipulation, requiring
no expensive apparatus." At the Bureau of Chemistry it was found that when
large amounts of copper were present the end reaction was occasionally indistinct,
otherwise the results were very satisfactory. The adoption of the method by the
association is reconmiended provisionally.
Another error in the determination of reducing sugars by the customary methods
is that resulting from the inversion of sucrose. This is particularly true of raw sugars
in which the inversion of sucrose is always a disturbing factor. This inversion is espe-
cially pronounced in the methods of copper reduction w^hich require long heating as
those of Defren and 0' Sullivan. By the latter method one of the cooperating chem-
ists obtained 8.67 per cent of reducing sugars on the raw sugar, a result far in excess
of that obtained by the other chemists using Allihn's method. The results by the
latter method are also manifestly too high. Mr. P. H. Walker by the method of
Munson and Walker obtained 6.56 per cent of reducing sugar as invert upon the sugar
sample in question. The referee obtained 6.69 per cent by his method when a cor-
rection factor is used, and Mr. Sherwood obtained 6.74 per cent by the process of
Meissl and Hiller — results closely agreeing yet considerably lower than the figures
obtained by Allihn's method. Mr. Halligan states, as a result of his experience,
that the Soxhlet method gives uniformly lower results than the method of Allihn,
and that he prefers the volumetric method on the whole both for speed and accuracy.
In the volumetric method when the solution is run in rapidly until nearly the point
of complete reduction, the error from inversion of sucrose is considerably though by
no means completely lessened.
Determination of Sucrose by Optical Methods.
The results olitained l)y the different chemists for sucrose by the optical method
are given in Ta])le 5.
122
Table 5. — Optical determination of sucrose in massecuite. sugar, and molasses.
[All readings corrected to normal weight. 100 cc. 200 mm tube.]
OFFICIAL METHOD.a
Massecuite.
Sugar.
Molasses.
Analyst.
Polarization. ' „
Su-
Polarization.
Su-
crose.
Polarization.
Su-
Direct.
Invert. ^™^«-
Direct. Invert.
Direct.
Invert.
crose.
H. P. Agee, sugar experi-
ment station, New
Orleans, La
Imposs
ihlf> to rend
30°
-1-81.11 —'^4.8(1
Per ci.
82.47
82.57
82.48
Impossible to read.
Do-
J. A. Hall. jr.. sugar experi-
ment station. New Or-
leans, La
do
-1-80.80
4-80. 71
31°
-25.30
32°
-24.86
J. E. Halligan, agricultural
experiment station, Baton
Eouge, La
do
Do
DILUTION METHOD,^ SAWYER.
Massecuite.
Sugar.
Molasses.
Analyst.
Polarization. ' „
Su-
Polarization.
Su-
crose.
Polarization.
Su-
.
Direct. ' Invert. <^^°^^-
Direct. Invert.
Direct. Invert.
crose.
H. P. Agee, sugar experi-
ment station. New Or-
leans. La '-I-36. 00
J. A. Hall, jr., sugar experi- I
ment station. New Or-
leans, La +36. 00
J. E. Halligan, agricultural i
experiment station, Baton
Rouge, La ' -h36. 80
29°
-17.16
28°
-17.60
30°
■17.16
Per ct.
41.28 -FS1.14
41.23 +80.86
41.83 +80.70
29° ' Per ct.
25.30 ' 82.74
32°
24.75
44 +22.52
39 +23.04
22° Per ct.
■15.40 i 29.51
31°
-14.52 { 28.82
32°
■15.71
DILUTION METHOD, c GEERLIGS.
Analvst.
Massecuite.
Sugar.
Polarization.
Su-
Polarization. ' ^ Polarization.
Su- Su- : ou-
Direct. I Invert. ^^°^^- | Direct. ] Invert.- ^^«^^- Direct. Invert. ^^^««'
H. P. Agee, sugar experi- '
ment station, New Or- | 30°
leans. La j+36.24 -17.6C
J. A. HaU, jr., sugar experi- i
ment station. New Or- ' 28°
leans. La +35.40 -17. 8e
J. E. Halligan, agricultural ■
experiment station, Baton I 32°
Rouge, La +36. 70 -17. 6C
C. A. Browne, jr.. Bureau of (
Chemistry, Washington, 26°
D.C +39.28 -20.4
Per ct.
41.89
30°
+81.24 -25.30
32°
+81.00 -24.57
32°
42.42 +80.88 -24.20
d46.02 i+79.52
I
24°
-28. 12
Per ct. ! 22° Per ct.
82.83 +22.76 -15.58 29.77
31'
^2.47 +22.80 -14.
30°
29.39
$2.03 +23.00 -15.84 30.34
82.38 +19.04 1-20.24 30.29
a 26.048 grams to 100 cc.
b 26.048 grams to 200 cc.
c 100 grams substance + 300 grams water. Normal weight of solution to 100 cc.
d High result due to leakage of molasses from sample during shipment.
All of the chemists reported inability to secure polariscope readings upon the
massecuite and molasses by the present official method, requiring the use of a normal
weight of material to 100 cc owing to the A'ery dark color of the solution. This method,
so far as low-grade sugar-cane products are concerned, has always been a dead letter
with chemists, and a change as to the manner of preparing solutions for polarization
is most desirable. Sugar chemists for many years past in their analysis of dark-colored
products have employed some method of dilution, either diluting from 50 to 100
123
grams of material with a known amount of water and making all analyses npon the
resulting solution, as advocated by Prinsen-Geerligs, or diluting a normal weight of
products to 200 cc, 400 cc, or 500 cc, as proposed by Mr. Sawyer in his previous
reports as associate referee on molasses methods. When a large amount of material
is available the former method is usually preferred, owing to the opportunity of secur-
ing greater uniformity of sample, a matter of no small importance in the analysis of
nonhomogeneous materials such as massecuite. The results of the chemists who
compared the two methods by dilution show no appreciable difference, and it is
recommended that these methods be adopted provisionally by the association.
An important question in connection with the polarization of low-grade sugar prod-
ucts is the lead precipitate error and what correction, if any, should be made for it.
This subject has aroused a great deal of discussion of late among sugar chemists in all
parts of the world ; but the various arguments pro and con are too well known to require
reviewing.
The influence of the strongly basic lead subacetate solution in precipitating reducing
sugars has already been referred to, and this would naturally have the effect of increas-
ing the rotation in the direct polarization. With low-grade sugar-cane products the
lead error from this source is unquestionably far greater than that due to volume of
precipitate. This error, however, applies only to the single and not to the double
polarizations. It will be noted in Table 5 that the single polarizations upon the sugar
and molasses show a wide variation between the results obtained in Louisiana and in
Washington. In his own work the referee employed but a small amount of lead solution,
and his direct polarizations are over 1 per cent lower with the sugar and nearly 3 per
cent lower with the molasses. With double polarizations, however, the results are
brought into very close agreement with those obtained by the Louisiana chemists.
A double polarization of raw cane products, whether of molasses, massecuites, or
sugars, is manifestly more accurate when it comes to the question of true sucrose con-
tent. Only in very exceptional cases, such as those noted by Geerligs a in the analysis
of Java products, resulting from strongly alkaline clarifications, do the reducing
sugars have an optical rotation of approximately zero, in which case alone would single
and double polarization give identical results. As a result of many comparative
analyses made upon low-grade sugars from Louisiana and Cuba, the referee found that
double polarization always gave higher results, the differences ranging from 0.60 per
cent in the case of a sugar polarizing 89.80° to 3.88 per cent with a sugar polarizing
72.40°. The optical power of the reducing sugars in these samples varied from —20.6°
to —46.9°, the general average being —34.5°. The theoretical value for the optical
power of a normal weight of invert sugar in acid solutions, according to Clerget's for-
mula, is —32.3° at 20° C. In other words, 1 part of invert sugar would neutralize
0.323 parts of sucrose at this temperature. W^iechmann b suggests the multiplication
of the percentage of reducing sugars by the factor 0.34 and the addition of this result
to the single polarization (17.5° C.) to obtain the true percentage of sucrose in raw
sugars. The individual variations in the optical power of the reducing sugars in^
different samples are so great, however, that the use of such a factor can not be
recommended.
Iij certain kinds of commercial work it is impossible from lack of time to make the
double polarization, and for purposes of control in sugar-houses a single polarization,
with proper regard to its limitations, is often sufficiently accurate.
The referee has spent a part of his time during the past year in search of a method
which would give the true direct polarization, and for this purpose has compared the
ordinary lead subacetate solution, Home's dry subacetate method. Pellet's method,
using hypochlorite of lime or soda and normal, lead -acetate solution, and a method in
which no lead at all was used, employing a little alumina cream and a new, powerful
a International Sugar Journal, 1906, p. 88. b Sugar Analysis, p. 99,
124
bleaching agent, sodium hydrosulphite. The results of this investigation, as far as it
has been carried, show that the highest polarizations are always secured by the use of
the subacetate solution, this being due both to the precipitations of reducing sugars from
solution as well as to the volume of the precipitate. The diy subacetate method gave
uniformly lower polarizations than the wet subacetate method, yet with this method
there seemed to be also in some molasses a partial precipitation of reducing sugars.
Pellet's method also gave lower polarizations than the wet subacetate method. The
lowest polarizations were uniformly obtained when no lead at all was used, the solu-
tions being simply bleached without any precipitation of sugars or interference by the
volume of the precipitate. The sodium hydrosulphite is entirely without action upon
the sugars, and the compound will be found of great service in the polarization of dark-
colored products.
Determixatiox of Sucrose by Chemical Methods.
The determination of sucrose in the three samples which were sent out by chemical
methods was reported upon by a number of the chemists, but the results were not com-
parable on account of the different methods and calculation tables used. The referee
has found a very general disagreement among chemists as to the proper interpretation of
the official method for this determination. The method (Bui. 46, p. 39) reads as follows:
"Determine first any reducing sugar in the sample, then invert the sucrose and redeter-
mine the reducing sugar. Deduct the percentage of reducing sugar obtained at first
and the remainder will be the reducing sugar derived from the sucrose . This multiplied
by 0.95 will give the percentage of sucrose. ' ' The disagreement among chemists comes
from the interpretation of the expression "reducing sugars." A number of analysts
determine the sugar before and after inversion as dextrose, a and multiply the difference
by the factor 0.95. Such a course is manifestly incorrect, since the factor 0.95 applies
only to invert sugar. Other chemists determine the sugar before inversion as dextrose,
levulose, maltose, lactose, or any other reducing sugar as the case may be, and that after
inversion as invert sugar, usually according to different methods. This course is even
worse than the former, since the reducing sugars present with the sucrose have been
determined before and after inversion by entirely different methods and tables. As
the official methods stands there is only one course to follow, and that is to determine
the sugars both before and after inversion as invert sugar by the same method and table.
The difference between these two results multiplied by 0.95 will give the true per-
centage of sucrose. The objection is raised that it would be incoiTect to estimate the
sugar before inversion as invert sugar, when sometimes it does not occur as such, as for
example in a condensed milk. This objection, however, has no force when dealing with
the sucrose determination. The object is simply to establish a zero point, so to speak,
upon the invert sugar scale, and it makes no difference what the reducing sugar is, pro-
vided it does not suffer hydrolysis or destruction in the process of inversion. After the
sucrose has been thus estimated the reducing sugar obtained before inversion may be
changed to levulose, lactose, maltose, or any other reducing sugar, either singly or sev-
erally, by means of suitable conversion factors, in accordance with the principles given
in the report last year upon the analysis of sugar mixtures. ^ In the particular instance
of sucrose and any single reducing sugar the table of Munson and Walker may be used
to great advantage ;c the weight of copper from the reducing sugar before- inversion
will give on the same line the zero point on the invert sugar scale as well as the corre-
sponding amount of dextrose, lactose, or maltose. The important point is that in work-
ing upon the analysis of sugar mixtures by reduction niethods a strictly uniform method
of procedure must be followed throughout.
« This error in procedure appears in Bulletin 46, page 24, section 8(b), in the method
for the determination of sucrose in cattle foods.
& J. Amer. Chem. Soc, 1906, 28: 439.
c Ibid., 1906, ^^.- 663; 1907, i?.9.- 541.
125
As a result of the work comprised in the foregoing report certain
recommendations were offered which are given on page 154, together
with the action taken thereon.
EEPOET ON CHEMICAL METHODS OF SUGAE ANALYSIS.
By L. S. MuNSON, Associate Referee.
The work of the past year on chemical methods of sugar analysis has been a con-
tinuation of that conducted for several years to establish uniform methods for the
determination of the various reducing sugars. It is proposed that the same method
of manipulation shall be used on all reducing sugars. Working along this line, tables
have been prepared by Mr. Walker and the referee for pure dextrose, pure invert
sugar, and mixtures of invert sugar and sucrose, and by Mr. Walker for pure maltose
and lactose. The preparation of the solutions, the method of manipulation, analytical
data, and tables for dextrose, invert sugar, and mixtures of invert sugar and sucrose
have been presented in the Journal of the American Chemical Society « in detail.
The results obtained upon lactose and maltose by Mr. Walker, together with full
details as to the work, may be found in the same journal under the title of The Unifi-
cation of Reducing Sugar Methods. ^ The two tables found in these articles give the
necessary data for determining all of the more common reducing sugars.
Mr. Davidson, chairman of the committee appointed to invite the
Secretary and Assistant Secretary of Agriculture to address the asso-
ciation, reported that the Secretary was out of town, but that the
Assistant Secretary would address the convention at 10 o'clock on
Friday morning.
The meeting adjourned.
THURSDAY— AFTERNOON SESSION.
THE CHEMICAL DETEEMINATION OP SULPHITES IN SUGAE PEODUCTS.
By W. D. HoRNE.
The official methods of analysis include one for determining sulphites (SO2) in
wine by distillinj,- with phosphoric acid (H3PO4) in an atmosphere of carbon dioxid,
catching the distillate in standard iodin and titrating with sodium thiosulphate. As
this method is seriously faulty, it is desired to call attention to the weak points and
to suggest certain improvements with a view to revision.
In such a residual mother liquor as a refinery molasses there are apt to be organic
substances which will volatilize from a boiling acid solution and be capable of reducing
iodin. Therefore, instead of the above indirect method it would be much better to
add hydrochloric acid and barium chlorid to the iodin solution and thus precipitate
the sulphuric acid formed by the oxidation of any sulphites (SO2) that have come
over.
Further, some suitable precaution should be taken against error due to the evolution
of hydrogen sulphid by the phosphoric acid (H3PO4) from any sulphids in the solution.
That sulphids do occur in appreciable quantities in residual sirups has not yet been
definitely settled, l)ut that they may occur is possible from the fact that boneblack
«June, 1906, 28: 663. & April, 1907, 29: 541.
126
used in refining sugar contains a little calcium sulphate -which is slightly reduced by
the carbon of the boneblack dimng revivication in the kilns to calcium siilphid,
and traces of this are at times given up to the sugar solutions flowing from the bone-
black.
The most promising method of separating hydrogen sulphid from sulphur dioxid
in the distillate is to pa^s the distillate through a Woulff bottle containing a 2 per
cent neutral solution of cadmium chlorid. which causes the immediate precipitation
of hydrogen sulphid as cadmium sulphid, while the stilphur dioxid is not precipitated
but passes on to the iodin. where it is oxidized to sulphtiric acid, to be afterwards
precipitated as barium sulphate.
As some sulphur dioxid may remain in the cadmium chlorid solution, it is simply
necessary to filter the contents of the Wotilff bottle into the iodin solution, thus sepa-
rating any cadmium sulphid and introducing the sttlphtu-ous acid into the iodin.
where it will be oxidized to sulphuric acid. The cadmium chlorid solution is without
effect on either the iodin or the barium chlorid.
By this method one can avoid reporting stilphids and organic reducing substances
as sulphur dioxid. Sulphid and sulphite if existing simultaneously in even moderate
quantities will be somewhat decomposed when heated in the presence of phosphoric
acid (HgPO^) and sulphur will be precipitated. It may be that they can not be
giA'en off together in appreciable quantities, but at any rate a preliminary test should
be made for stilphids before proceeding with any quantitative determination. This
can be conveniently done by acidif^'ing the diluted and warmed su"up with phosphoric
or hydrochloric acid, hanging a strip of filter paper moistened with lead acetate in
the test tube and bubbling carbon dioxid into the bottom of the tube. If sulphids
are present they will yield hydrogen sulphid. which will soon color the lead paper.
The above method is offered tentatively to the association with the hope that it, or
some better method, may be adopted in lieu of the present unsatisfactory mode of
analysis.
Mr. Wiley. The problem discussed by Mr. Home is of special
interest, both from a legal and a scientific point of view, at this time,
inasmuch as under the food and drugs act the determination of
sulphurous acid in foods in general must be made, and errors in the
methods for making such determinations call for special study.
For example, the addition of mineral substances to confections is
forbidden absolutely, not in large or small quantities, and, as sulphur
dioxid is a mineral substance, its determination in such products
entering interstate commerce becomes of imj)ortance. The mere
trace of a substance, wliich possibly is present naturally, ought not
to be considered as satisfactory evidence that it has been added, and,
as the presence of sulphids in food products may result from the
sulphur naturally present, the analyst should be able to discriminate
clearly between a natural product in a confection or other food and
an added substance: hence the importance of distinguisliing between
a sulphite and a sulphid, as pointed out by Mr. Home. The unre-
stricted use of sulphur dioxid in food products is, however, far more
£:eneral than it should be. For example, in sulphuring ^^'ine casks,
the wine may be racked four or -^xe times the first year, and possibly
fifteen or twenty times before it is completely ripened, and, if at
each racldng sulphur candles of any desired size are burned, the
127
total amount of sulphur deposited is excessive. Already steps are
being taken to prevent such excesses by fixing the length of candle
to be used. Dried fruit is also sulphured to improve its appearance
and protect it from insects; also, if sulphur is used, the evaporation
need not be carried so far, and thus more water may be retained in
the finished product. For these three reasons the sulphuring is
often excessive, and the product is sometimes resulphured before
packing. Apart from the question whether sulphur should be used
at all in foods there is no doubt but what such excessive use should
be prohibited, and the methods for accurately determining the
amounts present are therefore of increasing interest and Mr. Home's
paper deserving of special attention.
EEPOET OIT MEDICINAL PLANTS AND DEUGS.
By L. F. Kebler, Referee.
The passage of the food and drugs act, June 30, 1906, is undoubtedly the most
important occurrence affecting the status of drug products in this country since the
last annual meeting of the association. As the standards prescribed by the United
States Pharmacopoeia are by this enactment made the basis for determining the
quality of many drugs, and as future State legislation will undoubtedly be modeled
on the Federal law prescribing the United States Pharmacopoeia as the authority, the
character and reliability of these analytical methods assume increasing importance.
As a matter of fact, the Federal law with slight modifications has already been placed
on the statute books of Georgia and Louisiana and adopted as the health code of New
York City. In anticipation of governmental enforcement of the legal standards
considerable criticism of the methods has appeared, and their merits will undoubtedly
be thoroughly tested in connection with the administration of the law.
In designating the United States Pharmacopoeia and National Formulary, official
at the time of investigation, as the standards by which the law is in part to be inter-
preted. Congress apparently left open a way by which these standards may be modified
by action of the organizations which revise and publish them. As preliminary steps
toward such a modification, by means of a supplementary revision of the Pharma-
copcpia, have already been taken, the legal status of any alterations may have to be
adjudicated. In this interesting and unprecedented situation the study of drugs must
receive increased attention. State officials charged with the enforcement of the laws
governing the adulteration and misbranding of foods and drugs must turn their atten-
tion to drugs as well as foods. Cooperative study of assay methods, if participated in
and supported by those most directly interested in an equitable and effective admin-
istration of the law as applied to drugs, will result in very valuable contributions.
There is an element of truth in the idea that the assaying of drugs is beset with
inherent difficulties and peculiar sources of error, and that in its present stage of
development its results represent approximations to the truth in the interpretation of
which some latitude must be allowed. This idea is especially urged in anticipation
of the enforcement of the new legal standards of quality and purity.
The development of a high degree of accuracy in drug assaying is to be sought
through an application of the same scientific principles which have improved other
branches of analysis. The first necessity is a recognition of the nature of the diffi-
culties to be surmounted. The obscure causes of variation in results must be deter-
mined and eliminated. They should not be tolerated and retained in methods under
128
the euphemism "personal equation," or "personal error." A real personal factor is
obviously involved in using any of the senses as the standard of judgment, as sight
in determining the end reaction of a titration, but unless the details can be so arranged
as to reduce such personal equations to small proportions the method should not be
considered satisfactory in principle nor infallible in the administration of law.
If the large differences in the results by certain methods reported by collaborators
this year were entirely due to the personal factor in the strict sense of the term, they
would furnish evidence of serious defects in the methods. In so far as they originated
in lack of uniformity in manipulation, the obvious remedy is to determine the influence
of such details and to embody in the directions explicit requirements regarding every
detail which is found to be a material factor. The discretion which the methods allow
in greater or less degi'ee to the individual analyst must be intelligently and systemat-
ically curtailed. Only by such a process of elimination can the adaptability of the
methods be ascertained and the issue narrowed to features which may be radically
wrong or insusceptible of control. This, which is one of the principal objects of
cooperative work, will be most effectively promoted if collgiborators report not merely
quantitative results, but also criticisms and suggestions. If exact compliance with
the methods seems impracticable, the modifications adopted, and the reasons therefor,
should be fully detailed. Ambiguity or indefiniteness in the directions as well as
improvements in the manipulation should be indicated.
During the past year work on the methods for determining morphine in opium was
continued and methods for ascertaining all or part of the alkaloidal constituents of
cinchona, ipecac, and iiux vomica were added.
Samples of these drugs, delivered as being of U. S. P. quality, together with the
best available methods, were supplied to a number of chemists who signified a willing-
ness to participate in the work, and complete or partial results were reported by ten
collaborators.
The following directions were sent out with instructions that all calculations and
solutions be based on the data contained in the United States Pharmacopoeia, Eighth
Revision:
POWDERED OPIUM.
Method I — United States Pharmacopoeia, Eighth Revision, « with Additions.
Pun duplicates on opium as received.
1. Weigh the crude crystallized morphine on watch glasses, as directed by the
Pharmacopoeia, and report weights.
Mix and powder the morphine of the duplicates and test mixture as follows:
2. Determine the purity by the limewater method of the Pharmacopoeia, using 1
gram of the mixed crude morphine.
Report per cent purity.
3. "Mallinckrodt reassay. " — Place 1.2 grams mixed crude morphine in an 80 cc
Erlenmeyer flask, add 0.5 gram freshly slacked lime and 20 cc water, cork, and shake
occasionally for one-half hour. Filter into a similar tared flask with gentle suction
(reenforcing the point of the filter with a platinum or hardened paper cone), wash the
flask and residue with limewater until the total filtrate and washings amount to 35
grams. Add 3 cc alcohol. 20 cc ether, rotate, add 0.5 gram ammonium chlorid, cork
and shake vigorously.
Let stand two hours, then filter, dry, and weigh the precipitated morphine accord-
ing to the directions of the Pharmacopoeia. Report the weights obtained.
Method II — United States Pharmacopaia. Eighth Revision, Modified by Lamar.
Run in duplicate. Proceed as directed by Pharmacopoeia to precipitation of mor-
phine. To the 20 grams of aqueous extract add 60 grams of alcohol, stopple flask,
shake well for one minute and set aside for thirty minutes, during which time the pre-
cipitated material should have completely subsided. Decant the clear supernatant
liquid into a tared 250 cc evaporating dish, transfer the precipitate to a 7 cm filter pre-
"Abbreviated title "U. S. P. VIII."
129
viously moistened with a mixture of alcohol (3 parts) and water (1 part). The last
portions of the residue arc transferred to the filter by using small portions of the above
hydro-alcoholic solution. The filtrate is to be collected in the tared evaporating dish.
Continue washing the residue and filter by dropping the alcoholic solution on the filter
and the residue until the filtrate is no longer bitter. Add 35 cc of water to the contents
of the evaporating dish and evaporate on water bath to 14 grams, then proceed as
directed by the Pharmacopoeia.
Procure the same data for the morphine thus obtained as outlined under Method I.
Method III— Combination method.
Run in duplicate.
Macerate 12 grams of opium with water and proceed as directed by the U. S. P.
VIII to 20 grams of aqueous extract in tared flask, using, however, sufficient rinsings
to bring the weight to 22 grams instead of 20. Add 12 grams alcohol, exactly O.G cc
ammonia water (10 per cent), and sufficient water to bring the total weight to just 36
grams. Rotate to mix and filter at once, cover to prevent evaporation, collect just
30 grams in a suitable tared precipitating flask. Add 25 cc ether, rotate, then add
3 cc ammonia water (10 per cent), stopper, shake vigorously for ten minutes, then
proceed according to the IJ. S. P. VIII.
Procure the same data for the morphine, as outlined under Method I.
CINCHONA.
Run duplicates by each method.
Method I — United States Pharmacopoeia VIII.
Determine total alkaloids and ether-soluble alkaloids and report results.
Method II — Total alkaloid.
Place in a 200 cc flask 2.5 grams of the cinchona with 5 cc of hydrochloric acid
(10 per cent) and 15 cc of water, and digest on the steam bath for ten minutes. When
cool, add 50 grams ether, 25 grams chloroform, shake, add 5 cc of sodium hydroxid
solution (15 per cent), and shake for ten minutes. Add 1.5 grams powdered tragacanth
and shake again. Filter off 60 grams through a funnel containing a pledget of
purified cotton into a clean tared flask.
Extract by shaking out with successive portions of 20, 10, and 10 cc hydrochloric
acid (1 per cent), or until no more alkaloid is yielded, as evidenced by the test with
'Mayer's solution applied to a drop. To the united acid extracts, collected in a sep-
aratory funnel, add 15 cc chloroform, shake, make slightly alkaline with ammonia
water, and shake vigorously. When the chloroform separates, draw it off through a
double filter into a tared 100 cc Erlenmeyer flask and repeat the operation twice with
10 cc chloroform; evaporate or distil off the chloroform in the flask, add 3 cc ether,
evaporate, and dry at 110° to constant weight. Report weights.
Method III — Total alkaloid.
Place in a 200 cc flask 5 grams cinchona bark, add a mixture of 75 cc of ether with
25 cc chloroform, stopper the flask, shake and allow to stand for a few minutes, then
add 5 cc of ammonia water, shake and allow to stand three hours, shaking at inter-
vals. Decant as much of the mixture.as possible into a suitable small percolator, the
neck of which is plugged with a pledget of cotton, and receive the percolate in a
separatory funnel containing 20 cc half-normal sulphuric acid, or sufficient to make the
liquid distinctly acid after shaking. Rinse the contents of the flask into the percolator
and continue the extraction with additional portions of the ether-chloroform mixture
until the percolate gives no alkaloidal reaction, when a few drops are evaporated,
taken up with acid, and tested with Mayer's reagent.
The separator is to be shaken vigorously for about one minute, the layers allowed
to separate and the aqueous portion filtered through a pledget of purified cotton into
another separator and the operation repeated witli 10 cc more half-normal acid and
then with 10 cc of water. To the combined acid filtrates in the second separator add
20 cc of the ether-chloroform mixture and slight excess of ammonia water. Shake out
with three successive portions of the mixture, or until no more alkaloid is extracted,
and collect the ether-chloroform solutions in a separator. Rinse with a few cubic
centimeters of water, discard the latter, and draw off the ether-chloroform layer into
31104— No. 105—07 9
130
a tared flask or beaker. Evapjrate slowly and carefully, to avoid spattering, to dn*-
ness on the water bath, add 3 cc ether to the residue, again eA-aporate to drA-ness. and
dry at 110° in an oven to constant weight. Eeport weights.
Test the degree of exhaustion of the mai-c. Boil in a flask for ten minutes with
50 cc normal hydrochloric acid, cool, transfer the mixture to a percolator, the neck
of which is pro^-ided with a pledget of cotton, collect percolate in a separator, rinse
the marc wiih three successive portions of 10 cc of water, collecting all in above sepa-
rator. Eender the acid solution alkaline with potassimn hydroxid solution and shake
out ^ith three successive portions of 10 cc of chloroform. Wash the combined chloro-
form solutions with a few cubic centimeters of water, discard the latter, and evaporate
the chloroform carefully to approximate dr^mess in a tared beaker on the water bath.
Add 3 cc ether, evaporate, dry at 110°. and report weight.
IPECAC.
Eun duplicates by each method.
Method I — United States Pharmacopaia Till.
>a') Give results by tin-ation.
{b) Einse the titration residue with water and ether into a separator, add slight
excess of ammonia water, and shake out with successive portions. 25-1-20+104-10 cc
of ether: at each separation draw off the aqueous layer to a second separator and
transfer the ethereal layer to a third separator. The combined ether solution is rinsed
with a few cubic centimeters of water, the latter discarded, the ether drawn into a
tared flask, evaporated at tempera tm-e not exceeding 60° C. and the residue dried to
constant weight in a desiccator. Eeport weights.
Method IL
Place in a 200 cc flask 6 grams finely jwwdered ipecac, 120 grams ether, 5 grams
ammonia water, cork, and shake often for one-half hour. Let settle and filter off
tjirough a cotton plug in the neck of a funnel 100 grams i corresponding to 5 gi-ams
of ipecac) into a tared flask. Evaporate or distill off the ether and dissolve the res-
idue in 5 cc alcohol, add 20 cc ether. 10 cc water. 3 drops hematoxylin solution and
titrate with tenth-normal sulphuric acid, shaking strongly after each addition tmtil
the -^-iolet color changes to red-brown. Then add 30 cc water and continue titration
to a lemon-yellow color. Eeport number of cubic centimeters of tenth-normal acid
required.
Method III.
Place 5 grams finely powdered ipecac in a flask with about 50 cc ether, add 5 cc
ammonia water, and shake frequently for an hour. Then decant into a small perco-
lator, the neck of which is provided with a cotton plug, and rinse the remaining
powder into the percolator with small ponions of ether. Continue the percolation
with ether until no more alkaloids are extracted: collect the ethereal percolate in a
separator. Extract the ethereal solution by shaking out successively with 20— 10 -f 10
cc of normal acM. Combine the acid solutions, make sliehtly alkalme with ammonia
water, and shake out mixture successively with 20+10+10 cc of ether. Combine
the ethereal solutions. linse with a few cubic centimeters of water, discard the latter,
evaporate the ether in a tared flask, dr\- to constant weight in a desiccator, and weigh.
Eeport weights.
XrX VOMICA.
Eun duplicates by each method.
Method I — Vnited States Pharmacopma Vlll.
In evaporating the chloroform solutions of the total alkaloids use a tared flask, and
when the chloroiorm is apparently expelled add 3 cc ether, evaporate. dr>- to constant
weight at 105°. and weigh total alkaloid before dissolving in acid.
Eeport (o) total alkaloid by -weight, (6) strj-chnine=
131
Method II.
Place in a 150 cc flask 6 grams powdered nux vomica, 40 grams chloroform, 80 grams
ether, cork, and shake frequently for one-half hour. Add 5 cc of (10 per cent) ammo-
nia water, cork, and shake frequently for two hours. Then decant 100 grams exactly
(equivalent to 5 grams of drug) into a flask, distil it to a residue of 10 grams, add 30 cc
ether, 5 cc alcohol, 5 drops hematoxylin solution and 10 cc water, and titrate with,
tenth-normal acid to production of a red-brown color. Add 30 cc water and titrate to
a lemon-yellow color as the end reaction. Report the number of cubic centimeters
of tenth-normal acid required.
Acidify the titration residue with sulphuric acid, rinse into a separator, and shake
out three times successively with small quantities of chloroform, discarding the
latter. Make alkaline with ammonia water and shake out successively with
20 + 10 + 10 cc chloroform, or until no more alkaloids are extracted. Combine the
chloroform solutions, rinse with a few cubic centimeters of water, discard the latter,
and evaporate the chloroform to approximate dryness in a tared flask on a water bath.
Add 3 cc ether to the residue, evaporate to dryness, and dry residue to constant weight
in a desiccator. Report as total alkaloids by weight.
Method III.
Place in a flask 10 grams of nux vomica powder, add 100 cc of a mixture of 75 cc
ether, 25 cc chloroform, 8 cc alcohol, and 3 cc stronger ammonia water. Cork tightly
and macerate with frequent shaking for twelve hours. Decant as much of the mix-
ture as possible into a suitable small percolator, the neck of which is plugged with a
pledget of cotton, and receive the percolate in a separator containing 25 cc normal
sulphuric acid, or sufhcient to make the liquid distinctly acid after shaking. Rinse
the contents of the flask into the percolator with a mixture of ether and chloroform
(3 to 1), pack the powder with a glass rod, and continue extraction with the same
mixture until the percolate gives no alkaloidal reaction when a few drops are evapo-
rated, taken up with acid and tested with Mayer's reagent. The separator is to be
shaken vigorously for about one minute, the layers allowed to separate, the aqueous
portion filtered through a pledget of purified cotton into another separator, and the
operation repeated successively with 20+20 cc more of half-normal acid and then
with 10 cc of water. Make the combined acid filtrates in the second separator alkaline
with ammonia and shake out successively with 25 + 15+15 cc of chloroform. Rinse
the combined chloroform solutions with a few cubic centimeters of water, discard the
latter, and evaporate the chloroform approximately to dryness in a tared flask or beaker
on the water bath, add 3 cc of ether, again evaporate, and dry the residue to constant
weight at 105°. Report as total alkaloids by weight.
The results submitted by the various cooperating analysts are given in the accom-
panying tables. In order to provide a common basis of comparison for the variability
in results, obtained by different methods from the several drugs, the ratio of the dif-
ference to the average is given in the form of a percentage variation in addition to the
average, maximum, minimum, and difference. It is to be noted, however, that these
summaries do not include certain widely variant results (indicated in the tables),
which apparently are notfairly representative. The significance of anomalous results
is recognized, but their interpretation involves special considerations. An expe-
rienced and observant analyst can often associate an abnormal result with some unu-,
sual phenomenon in the operation, which would naturally lead to a repetition of the
analysis. On the other hand, abnormalities may become apparent only upon com-
piling and comparing a number of results, as in this work, in which case it may be
difficult to assign any cause for them. Such abnormalities militate somewhat against
a method, even though the results are in the main satisfactory.
132
3C X X
o- o -CI t^ t^ o
x' ^" t^ ci X ^
M CT. -^
,1 = J >-.
X i?
s» — . = = = —
XXX
< ~
Si,
O
6
^•i>>
"? C 2
t^ C~. X
-^ iC X
i-1 5^ ^
5 n
X ci L-:
t- X r-:
■^'oi c4 -^ -^ Q
C:X ,— i ^
5 23 2 23
2 rca: S
= CO
:0 t--^'CC O
■o rrcc o
o ~ -^ -^
iX C2X —^ —
I i: I
t- ^x ^
. X =n
t^ 5 aT
— ^ tj
c .c f-
SSI
X t-~ js
X "?• r: — M
X X O O iOi.^»0:S -
c; X CO oi a \ c^c-i ^^ t^|_r
c c o
X X X p:
t^ X Xj'c:t-:j^ ?i
I
X r^ «-- — o X L":
X --;; t^ X O X rr
X - c: C-. X ~ M
e t- ^5 6i c; ^ I ^ r ?: ~
^■^~=?
^xx X XX — xir. X
rt — =; X ■
X d ^i i>:
cc X 'O — ! . r^
XX--
M C-1 C: — — Id
'So ^ ^
cod (N d
ix:ii: = ;^
C: X O O 2^3
I- .-: o — n t^ L- t- "3 1^ 2 -
M ri M —
, • ■ , ■ ] . . "t; . _ • 1 I ■ J _• !__;-,' • ^- ' ^ ' ' '
> . I
= ~ •> ~ > ~ -^ ■
X X 5 ^
§S ^
133
Opium.
In princii)l(> tlu^ three methods studied this year differ essentially only in the man-
ner and order of separating impurities. The original precipitate in Method I is as-.
sunied to be impure, the i)ortion thereof solul)le in limewater being accepted as the
equivalent of pure morphine. In Methods 11 and III preliminary purifications of the
solutions are undertaken to make practicable the precipitation of pure morphine,
thus obviating the necessity of further treatment. The additional tests directed in
this work are intended to measure the accuracy of these respective plans. The aver-
age per cent of limewater-purified morphine by Method I (10.76) should be compared
with the average per cents of morphine as precipitated by Methods II and III, which
are 10.87 and 11.10, respectively. The limewater-correction results by Methods II and
III are 10.50 per cent and 10.51 per cent, respectively.
From these data the original results by Method II are fairly in accord with the cor-
rected results by Method I, and the preference between the two methods is therefore a
question of expediency rather than of accuracy. The original results by Method III,
however, vary through a somewhat wider range and the percentage of impurity asso-
ciated with the morphine is materially higher. The inferiority of Method III in its
present form is therefore apparent. A comparison of the crude and the purified mor-
phine by Method I tends to confirm former observations, namely, that the impurity
varies practically independently of the amount of morphine. Stevens's results con-
firm our former observation, namely, that itris possible to precipitate a very pure mor-
phine and obtain an amount approximating closely the best results. Schulz's results
are significant. It will be noted that he obtained considerably the highest corrected
result, and next to Stevens the highest purity, by shaking thirty minutes and precipi-
tating five and six hours. The influence of the time and manner of shaking and the
temperature and period of precipitation upon the amount of original precipitation in
Method I has often been pointed out, and it is not improbable that these factors affect
the morphine as well as the impurities. A more detailed study of these points seems
advisable.
Method III is an adaptation of the principle of purification by partial precipitation
with ammonium hydroxid, introduced by Dieterich, and embodied in the method of
the German Pharmacopoeia, third edition. The aliquot feature is replaced by total
extraction of the opium. The precipitation is made in presence of alcohol, as in
Method I. The preliminary results obtained in the drug laboratory by this method
were encouraging. Some collaborators, however, experienced difficulty in obtaining
the 30 grams of filtrate after addition of ammonium hydroxid, and the slow filtration
involved exposure to evaporation. The operation might probably be expedited by
providing a larger amount of opium and solvents from which to filter 30 grams. But
the olijection seems well taken that the sarne amount of ammonium hydroxid would
not be suitable for different opiums. On the whole the method does not seem adapted
to produce as pure morphine as Lamar's (Method II).
As stated in the report of this work for 1905, the experience of two years, shows that
nothing is gained by removing the morphine from the counterpoised filters for weigh-
ing and the results are 0.1 to 0.2 per cent lower. Further study of this point seemed
needless, and weighing on watch glasses was directed simply because it is the phar-
macopoeial method.
Limewater undoubtedly dissolves some of the impurities precipitated with the
morphine, and the purity factor thus obtained is too high. Lyons believes that time
is a factor in this solubility, and that a portion of the impurity first dissolved will repre-
cipitate on standing. The color alone is ocular evidence of soluble impurity in the
limewater solution, and according to present theories the calcium ammonium meconate
of the original precipitate is l^y limewater changed to calcium meconate, which may
complicate matters.
134
Unsuccessful attempts have been made in the diiig laboratory to determine volu-
metrically the moiphine in the limewater filtrate. One method was to determine the
total alkalinity (^f the solution in comparison with a blank experiment with the same
limewater. the increased alkalinity being attributed to moi-phine in solution. The
other method was based upon the alleged acid or phenolic reaction of morphine toward
Ponrier's blue, according to which in parallel experiments the morphine should pro-
duce a decrease of alkalinity by neuti-alization of the lime. The experiments failed
for lack of definiteness in the end reactions. The addition of alcohol proved of no
advantage.
The Mallinckrodt reassay has been suggested as a useful check on the puiity of mor-
phine precipitates. It is not expeditious, and its results need correction by factors
not yet well determined. The need of an accm'ate (even though tedious and compli-
cated i assay process for opiiun as a standard of comparison for shorter technical meth-
ods is obvious, and in view of the recognized defects of the limewater correction it
seems expedient to attempt to control results by an additional method.
From considerable experience with the reassay method. \lallincki"odt l^elieves that
a con-ection of 20 to 30 milligrams should be added to the weight of reassay morphine
for solubility in the mother liquor, which would be equiA-alent to raising all results
about 0.21 per cent. A few blank experiments with practically pm'e morphine in the
drug laboratory indicated that the solubility at 25° to 28° is about 30 milligrams. On
the other hand Mallinckrodt' s experiments seem to show that the morphine obtained
by this method is only 97 to 99 per cent pm-e. Three or four reprecipitations seem to
be necessary to eliminate all the impurities. The application of a con-ection factor
should be subject to futm'e experiments.
Lyons questions the use of alcohol in the reassay, and also thinks less ether would be
preferable. He has obtained encom-aging results by the substitution of liquor calcis
saccharatus (^British Pharmacopoeia • for lime, both in the reassay and in opium assays,
but he especially recommends a trial of Fliickiger's suggestion, the preliminary
removal of calcium from the opium solution by means of ammonium oxalate. A pau'
of duplicates run by this method in the drug laboratory gave an average total of 11.85
per cent— con-ected by limewater, 10.94, and by reassay, 10.26 per cent.
Lyons also objects to the pharmacopoeial dnection that the precipitate l^e washed
with morphinated alcohol, as moderate changes in temperatm-e affect the solubility,
causing the solution to deposit or dissolve morphine, and also morphine is apt to be
deposited by evaporation. He prefers liberal washing with morphinated water, and
drying the filter between folds of absorbent paper.
ClXCHOXA.
The sample of cinchona bark sent out was below the standard, which is not less than
5 per cent of total alkaloids and at least 4 per cent of ether-soluble alkaloids. This
circumstance is favorable to the method, which might give low results with rich barks
owing to imperfect extraction. The fineness of the powder also favors complete extrac-
tion. None of the methods can. on the whole, be regarded as entirely satisfactory in
view of the results reported.
The results by Method I average 3.49 per cent, with a variation of 19.5 per cent. The
figures do not suggest any single or predominant cause of variation. Comparing them
with the results of Method III, which is the best one available, the average total, 3.49
per cent, by Method I. is in substantial agreement with the average total, 3.37 per cent,
by Method III. In fact, if the latter result be corrected by excluding the results of
two analysts the average totals for both methods are 3,49 per cent. This affords no
support to the criticism that aliquot methods like Method I are likely to yield high
results owing to the fact that the aliquot portion of the solvent contains more than the
theoretical fraction of the active constituents of the drug. The possibility of a com-
pensating loss elsewhere in the process must not be overlooked. In this connection it
135
may be noted that Method 1 assumes that 125 cc of ether plus 25 cc of chloroform makes
150 cc of ethereal mixture. Though the resulting condensation in volume may be
insufficient to lead to materially higher results, it would be better in principle to
measure 150 cc of a similar mixture after cooling.
Regarding possible causes of the differences in the results by Method I the following
suggestions received seem especially worthy of note. Lyons considers the amount of
acid directed to be used in shaking out the ethereal solvent insufficient to extract the
latter thoroughly. This would make both the total alkaloid and the ether-soluble
alkaloid low, other considerations being equal. Parker questions the suitability of
the graduated cylinder for dividing the 50 cc of acid solution into equal parts. An
error in the division would give a total alkaloid gain at the expense of the ether soluble
or vice versa. Inspection of the results reveals examples apparently sujDporting both
criticisms, but the extreme variations in the ether-soluble results are probably due in
great degree to other causes.
Method II is Fromme's method. « The results by this method would, as a whole,
appear to be extremely unfavorable, with the low average total of 3.10 per cent and the
very high percentage variation of 60.9. Inspection shows that the results may be
separated into two groups, one containing the 8 results of Dohme, La Wall, Parker, and
Schulz with an average of 3.68, maximum 3.94, minimum 3.45, and percentage varia-
tion of 13; the other separated by the marked interval of 0.87 per cent, containing 6
results by the remaining three analysts witn an average of 2.33, maximum 2.58, mini-
mum 2.05, and percentage variation of about 23. Wetterstroem's result of 1.68 is
excluded as anomalous.
It seems reasonable to conjecture that this peculiarity is due to some single difference
in the manipulation which might be revealed if a careful comparison of notes were
possible, and which might place the method on a better basis. One of the principal
criticisms made by Asher and Lyons, who obtained low results by Method II, is that
the maceration period (ten minutes) is too short. It would seem that the preliminary
digestion with hydrochloric acid justifies some shortening of the time of maceration.
Dohme attributes his higher results by Method II to the larger proportion of solvent
to drug than was the case in Method I, causing a more complete extraction. Such a
difference might be more apparent with a richer bark. Parker suggests that the varia-
tions might be less (though more time might be necessary) if the quantity of bark were
doubled and the other quantities increased in proportion; also that the filtration of
alkaloidal solutions, especially as applied in this case, through double filters, is an
operation to be regarded with suspicion in view of the imperfectly known phenomena
of adsorption and the doubt of thorough washing. Method III, as a simple total
extraction method, was intended to furnish a check upon the preceding two. The
results are in a measure disappointing, the percentage variation being 22. Analysis
of the results show that they fall into groups, not so marked but similar to those dis-
cussed in connection with Method II. If the considerably lower results of Asher and
Lyons be set aside, the average total of the remaining results is 3.49 per cent and the
percentage variation only 11. That the variability is partly due to imperfect extrac-
tion is indicated by the fact that all the low results correspond to high yields of alka-
loid from the marc. The amounts of alkaloids extracted from the marc show plainly
that failure of the last portion of ethereal percolate to respond to the test for alkaloid
with Mayer's reagent is no proof of exhaustion of the drug, and that additional treat-
ment is advisable. Preliminary digestion with acid (analagous to Method II) with
subsequent drying is suggested. Lyons points out that the alkaloid extracted from
the marc in the manner directed is impure and that purification by re-solution in acid
and shaking out again may materially reduce the amount.
o Csesauand Loretz, Geschafts-Bericht, 1905, p. 44.
136
>
-1
^~
t M r^ rt p-
c-ic
g5
ce
•ceo
CON
^
8 1
i
iS
gag
So
32
ccco
O
CO
28
1 , = rssx^
^ - i « ^j :^' j^
f3g
8S
rcce
1 ^
1 c^
cecsj ! "£2
i
g2 J?5^^
r-^ — :j"^i rQ — —
s s
■^«
8g
re
£:2
S
^i ?i ?< H cc ^; re
cccc
ece4
'ci
1^ ?^-| , f"
5; 5t
§gs
<BS
^
1= Slll ^
33
gg
:
^J5
?5
-
---^
m
w
gg2
i
.^1 1"""" = """
3 S X i 3 -2
r
§3 i«^
i ^
i g
i
1
^ If ii---:
-2;
s
?5 = ^- S
< 1 |!§3J351S
gs
-^
^«
1 "
I ;
5 ^i i^?^^izs^^^
^— ■^— i^ ^i ri re rt r-f H c^i ^^
5s
-5g§
s g s
TO
coco
CO \
:i
^c |:^H^-5gs^
^— -
:^i ^5 ?i T^i ri r^ ^i ^i i-i ^J
2=>
WO)
i
1 -3 1^^--^--^
rcro
cc « rt re re r^ re r'i eo
coco
§ue
1
■
>>
1
1
<
•r
i
>
i-
•—
X
1
5
1
1
X
>
;
^
^
•r
; i
; j
i 1
>
5:
^1
5g
-=2
^11
so ■>•■««
137
_5
•^
bio be
">^
Tj
§
X3
j-T
^1
S|
<»
^
-§
'~!S
0 q;
S
+^'3
, H
>>
bo
§1
SI
1
a 5
s
C
_>
■5
1
DO
§■
'>.
0
tt
cl 0
■3
>
1
S
s
T)
-c^
s
t^
m
« c3
o
C
1*
m
tn
-3 '3
0.2
O'i
0
^
.S-o
C C
TJ
c
all
•* 0 en
^1^
•" 0
05
.s§s
la
^Jrt Q> 3 3 a- cS
,c ®
„ c ^ ^ .2 .5P P =s
„ J-. be-- M bjo-u G -o
c o o .2 r^ o 'E "S c t^ S
"^ o R c 0^ C C
<S «« S C-^
.3 O c'
.0 OS &
ce « ^
1 !-c ^ ai 02 CO
138
The results on ether-soluble alkaloids by Method I vary over a range of 36 per cent,
which is unsatisfactory. The average is 2.72, maximum 3.14, and minimum 2.16
per cent. Parker's results alone indicate the cause of the differences. Criticisms
received and observations made in the drug laboratory point strongly toward indefi-
niteness regarding temperature conditions and practical difficulties in carrying them
out. Lyons considers that it has been demonstrated that practically all the cinchonine
is separated by ten minutes continuous shaking at a temperature as low as 15°, and
obtained his results in that manner. By following the pharmacopoeial directions,
agitation will take place at a temperature below 20°, but not necessarily as low as 15°.
Two minutes shaking within this range of temperature often hardly starts crystalliza-
tion, which proceeds rapidly at a somewhat lower temperature. Though too few in
number to justify a definite conclusion, the results of Parker tend to support the view
that two minutes shaking is sufficient under proper temperature conditions, but that
the latter should be rigorously prescribed. To maintain a uniform temperature in
warm weather, during the shaking and standing period, is not an easy matter without
some special arrangements, and the improvising of these is }eft entirely to the analyst.
Unless the solution is chilled below 15° to begin with, the addition of the ammonia
water and the shaking are likely to raise the temperature above 20°. If a constant
temperature bath be used and the separator removed and shaken, the temperature of
the contents will rise in warm weather, and some arrangement for shaking in the bath
seems necessary. The ether used for rinsing should also have the same temperature
as the main portion. It is sometimes difficult to separate the ether solution clear, the
crystals remaining in suspension. After tapping off the aqueous layer the ether solu-
tion can be better decanted through the mouth of the separator, thus avoiding con-
tamination with portions of the aqueous layer adhering to the inside of the lower outlet.
The neck of the separator should finally be washed with ether.
In connection with both the total and ether-soluble determinations of Method I,
the comment is made that for gravimetric determinations, unless good reasons to the
contrary exist, the final ether-chloroform solutions (obtained by shaking the alkaloid
out of aqueous solutions) containing soluble salts (in this case ammonium sulphate)
should be rinsed with a small amount of water before evaporating. Otherwise the
ether may carry enough saline matter in solution to give high results. The mutual
solubilities of water, ether, and chloroform should not be ignored.
Ipecac.
If all results were included, pharmacopoeial Method I would appear at a disadvan-
tage with a 45 percentage variation. By eliminating the low results of one worker the
percentage variation is reduced to 21.3, which compares favorably with those of
Methods II, 22.5, and III, 24.2, Even by excluding these results it is apparent that
Methods I and II give lower results than Method III. Aliquot Methods like I and II
ha^.e been charged with a tendency to yield high results due to several inherent errors.
If such be the tendency it does not appear in these figures. Possibly the maceration
period of one-half hour is too short. The gravimetric modification of Method I
evidently gives more uniform results than the pharmacopoeial titration method.
In Method I the questionable principle is adopted of assuming that 115 cc of ether
and 35 cc of chloroform are equivalent to 150 ccof the mixture. The necessity of the
admixture of chloroform is not apparent and the use of ether alone, with a smaller
proportion of drug as in Method II, might repay trial. The agglomeration of the
drug can not be produced by 10 cc of water, 18 to 20 cc being necessary. The shaking
out of the alkaloid from the alkaline solution before titration as directed is at times
incomplete.
Wetterstroem obtained no definite end reaction with hematoxylin in Method I and
used cochineal in all titrations. The experience of the drug laboratory with hema-
toxylin in Method I was very unsatisfactory. When used in the amount directed
139
(5 drops), no definite end reaction could be obtained with either freshly prepared
or old solutions of two different samples of the indicator, all of which in blank tests
proved sensitive to the titrating solution, fiftieth-normal potassium hydroxid. The
first apparent and best marked change was a fading away of the slightly yellowish
tint of the alkaloidal solution. On adding fiftieth-normal potassium hydroxid drop
by drop the solution very gradually assumed a smoky bluish-gray appearance, with
apparent separation of alkaloid. No purple or violet color was ol:)tained. With 30
drops of hematoxylin solution slightly better results were produced, the color with
excess of alkali being a pale smoky blue. The results of various experiments, which
it is impracticable to describe here, seemed to indicate that neither carbonic acid
nor overheating of the alkaloid is the disturbing factor, but that the alkaloidal sul-
phate interferes with the color reaction between the free alkaloid and hematoxylin,
and strengthened the impression that this indicator is not adapted to the titration of
ipecac alkaloids as directed in Method I. Cochineal gave a much more satisfactory
end reaction.
Method II is that of Panchaud. The titration to an acid reaction with hematoxylin
(an inversion of the usual method) is characteristic of Panchaud's scheme of analysis,
which will probably be adopted for a number of drugs in the forthcoming revision of
the Swiss Pharmacopoeia. It has the supposed merit of eliminating the usual back
titration with alkali.
In the drug laboratory the end reaction with hematoxylin in Method II, while not
entirely satisfactory, was considered less liable to cause large errors than the condi-
tions of Method I. Considering the simplicity of Method III, the results show sur-
prising variations. Possibly the directions do not insure thoroughness in the extrac-
tion and shaking out processes.
Nux Vomica.
By Method I the variation in total alkaloid was 19.5 per cent, or, excluding Lyons's
low results, only 9 per cent; in the strychnine titration, excluding Parker's two high
results, 30.1 per cent. In Method II the variations were, by titration 12.5 per cent, by
weight 9.2 per cent, and in Method III, 14 per cent.
In Method I the pharmacopoeial directions for measuring 200 cc of the prepared
ethereal mixture commendably depart from the principle criticised under ' ' Cinchona ' '
(I) and "Ipecac" (I). Particles of the drug are likely to float in the ethereal liquid,
and it would be advisable to decant the latter through a small percolator provided
with a pledget of cotton in the neck. The shaking out with normal sulphuric acid
as prescribed does not remove all the alkaloidal matter and should be repeated until
Mayer's solution gives no reaction. Traces of alkaloid were still extracted after
shaking out three times more with 5 cc of the acid.
Parker assayed the remaining ethereal solution and marc from the two total alkaloid
determinations under Method I, which yielded 3.02 and 3.04 per cent. The marc
was extracted successively with chloroform, hot dilute sulphuric acid, and alcohol,
the combined solutions concentrated and shaken out. By this procedure these
portions yielded 2.89 and 2.87 per cent, respectively, by weight. This would indicate
that the sample contained 2.95 per cent of total alkaloids, which is in close agreement
with the results by Method III and supports the theory that the aliquot methods give
slightly high results.
The results of the strychnine determinations are quite irregular. Experiments in
the drug laboratory readily located a source of error in the reaction between nitric
acid and brucine. It was repeatedly observed (about one time in four) that under
apparently identical and strictly pharmacopoeial conditions, the characteristic red
color developed sometimes promptly, sometimes slowly, and sometimes not at all,
even after hours. In one case the alkaloid was recovered, again subjected to the
nitrating process and recovered, and still gave a strong reaction for brucine.
140
Examination of the literature showed that the pharmacopoeial method has also given
erratic results in the hands of other analysts, and that the causative factors have not
been fully determined.
Howard," working with Keller's method, concluded that the temperature is an
important factor, and obtained the best results by nitrating at 0°.
Farr and Wright^ using the pharmacopoeial process foimd that only 9 per cent of the
brucine was destroyed at ''ordinary temperatures," and not quite all at 38° in ten
minutes or at 50° in five minutes, but that it was entirely destroyed at 50° in thirty
minutes. In their formal directions, however, they direct the use of 3 cc of nitric acid
sp. gr. 1.42 (instead of the TJ. S. P. sp. gr. 1.40) and nitrate at 50° for ten minutes.
Reynolds and Sutcliffe c found that dilute nitric acid fTce from nitrous has no
oxidizing action on brucine and that when the reaction failed, it could be started by
the addition of a minute amount of sodiimi nitrite — for example, 0.5 cc of 1/1,000
solution. When nitric acid of sp. gr. 1.42 was employed, the reaction never failed to
occur satisfactorily at ordinary temperatures from 14° to 25° in ten minutes, but higher
temperature or prolonged action is liable to cause decomposition and loss of strychnine.
Gordin's original method^ directs the use of nitric acid of sp. gr. 1.42. x\s the
method is incorporated in the Pharmacopoeia the official nitric acid (sp. gr. 1.40) is
specifically required. The difference of about 4 per cent in content of absolute nitric
acid may be immaterial, but if the presence of a certain amount of nitrous acid is
essential the directions require revision.
The Pharmacopoeia in the purity rubric of nitric acid does not include a test for
nitrous acid. Krauche mentions the reduction of potassium permanganate by nitric
acid as an indication of the presence of hj^onitric or nitrous acid and states that owing
to the tendency to 'decompose they are nearly always present in the strong acid.
According to Silberrad / strong solutions of nitric acid are only with difficulty freed
from traces of nitrous acid. Re^Tiolds and Sutcliffe were able by the use of urea and
barium or sodium peroxid to so completely free fairly strong solutions of nitric acid
from nitrous that they did not react with brucine, but never succeeded in doing this
with a 1.42 sp. gr. nitric acid. They also noted that the permanganate reaction is not
so delicate a test for nitrous acid in nitric as the brucine reaction.
It seems probable that nitric acid of sp. gr. 1.42 is more likely to contain nitrous acid
than the pharmacopoeial acid of sp. gr. 1.40, but until the invariable presence of nitrous
acid in the former shall be established it e\T.dently is advisable to direct the addition
of a proper amount of sodium nitrite in the determination of strychnine in nux vomica
by the nitric acid method.
The nitric acid employed in the experiments in the drug laboratory had a specific
gravity of 1.4041 at 20°, and at the time of its receipt, some three months before, had
failed to react for nitrous acid with the permanganate test. The irregularity of the
nitrating action with this acid suggested that some impurity might be responsible for
the slight reduction necessary to start the reaction, but no such agency was discovered.
The most probable impmity seemed to be traces of solvent remaining in the alkaloidal
residue.
The results by Method II (Panchaud's titration method) are reasonably concordant.
In computing the titration results the factor 0.03615 (the mean of brucine and strych-
nine) was used. But as the results by Method I show that the total alkaloidal matter
"Analyst, 1905, 30: 261.
^Pharm.J.,1906,77;83.
J. Soc. Chem. Ind., 1906, 25: 512.
dKxoh. Pharm., 1902, 240: 641.
cElrauch. Testing of Chemical Reagents, 3rd ed., p. 186. Translated by William-
son and Dupre.
/J. Soc. Chem. Ind., 1906. 25: 156.
141
contains about 40 per cent instead of 50 per cent of strychnine, the true factor is 0.03675
and the true average titration result 2.85 per cent, instead of 2.81 per cent total alkaloid.
In comparing the vohimetric with the gravimetric results the different directions
in j\I(>thod II for dryingthe total alkaloids before weighing should be noted. It would
perhaps be* unsafe to assume that drying to constant weight in a desiccator would
dehydrate the In-ucine as completely as drying to constant weight at 105°, which is
directed in Methods I and III. Schmidt « states that crystallized brucine containing
4 molecules of water loses part of its water of crystallization at ordinary temperatures
and all of it in vacuo over sulphuric acid or on heating to 100°.
The end reaction with hematoxylin as in "Ipecac" (II) was found not to be as sharp
as is desirable, though not considered seriously objectionable. Method III, like the
corresponding total extraction methods for cinchona and ipecac, gave too variable
results to be entirely satisfactory for purposes of control. Like cinchona, nux vomica
is a difficult drug'to exhaust, and it is possible that in some instances, notwithstanding
the test with Mayer's reagent, the marc was not entirely extracted. The thoroughness
of the shaking-out process was not directed to be tested, and in this slight losses may
haVe occurred.
Mr. Wiley. The subject of drugs will probabh^ be of increasing
interest to agricultural chemists in the future and especially to this
association. There is a decided tendency to modify State laws as
regards their general provisions in respect of foods and drugs to corre-
spond with the pro^dsionsof the National law. Tliiswill bring to the
State officials, who in many cases consider onl}- foods, the similar
work in regard to drugs, and as much of this work is done b}^ experi-
ment station chemists, the drug work will come largely to the same
analysts. Everything relating to drugs, therefore, and especially
methods for their examination, has a greater interest in this associa-
tion than ever before. The Bureau of Plant Industry also is endeav-
oring to introduce into this country a great many drug plants not
previously gro^^^l here, and in this wa}^ all chemical problems related
to the production and assa3dng of drugs will come more directly in
touch with the work of the agricultural chemist, whereas heretofore
they have been considered principally by the pharmaceutical chem-
ists. From both points of view, therefore, this section of the work is
of great and increasing importance.
^Ir. Kebler. Although some progress has been made in the culti-
vation of drug plants in this cojintr}^, it is virtual!}^ impossible to
compete with the imported products because of the low quality of the
imported goods and the fact that they are produced by very low-
priced labor. The food and drugs act, however, provides that if a
product is recognized in the Pharmacopoeia and does not comply with
the standard set by this authoritA", while it may be imported, it must
be marked to show its strength. This correct labeling of inferior
products will enal)le the first-class products to ccmipete on a fair
basis.
"Lehrbuch der Pharmaceutischen Chemie, Part 2, p. 1398.
142
It should be noted that the chairman of the revision committee of
the Pharmacopoeia intends to issue a supplement of changes embody-
ing corrections and defects in the standards that may be pointed out
to him, and while this will not invalidate the Pharmacopoeia as a
standard as far as the National law is concerned, it may have a dif-
ferent effect in the States, and it might be well for State chemists to
consider the matter and communicate with Professor Remington, of
Philadelphia, who is in charge of the revision.
EEPOET 01^ SOILS.
By J. H. Pettit, Referee.
In following out the instructions of the association on soil work this year, the referee
sent to each chemist expressing a willingness to take part four samples of soil. Of
these, sample No. 1 represents the level prairie land of the Lower Illinois glaciation
in the southern part of the State. It is a gray silt soil with a rather stiff subsoil and lies
too flat to give good surface drainage. Samples Nos. 2, 3, and 4 are from soil types
included in the early Wisconsin glaciation occupying the northeast quarter of the State.
No. 2 represents the gray silt soil of the timber land. No. 3 the brown silt soil of the
gently rolling prairie land which is naturally well surface drained, and No. 4 the black
clay loam which occupies the lower and poorly drained areas of the prairie and has
receiA^ed the wash of the higher, more rolling lands. Under the conditions found in
the field these soils vary in productive power in the order named, sample No. 1 being
the least productive, and not differing markedly from No. 2. In all fairness it must
be stated that the phosphorus content of sample No. 1 shows that it represents the
better phase of this type of soil, while that of sample No. 3 shows that it represents the
poorer phase of the brown silt loam. For all of these soils 320,000 pounds may be taken
as the weight of an acre-inch.
With these samples the following directions for analysis were sent out:
(a) Determine dry matter upon 5 grams of the air-dried soil by heating five hours in
water oven.
(6) Weigh 400 grams of the air-dried soil into a 2J-liter flask, add 2,000 cc of distilled
water, shake vigorously three minutes, let stand twenty minutes, and filter. Use an
unglazed porcelain filter « if possible. If this is not convenient, filter through paper,
shake up with 10 gi-ams of carbon black^ and refilter. Evaporate 1,500 cc of the clear
solution to 25 cc, add 2 cc of concentrated nitric acid, neutralize with ammonia, make
barely acid with nitric acid, add 15 cc of ammonium molybdate solution c and keep at
50° to 60° for two hours. Lefstand overnight, filter, wash with 0.1 per cent ammonium
nitrate solution until free of acid, and twice with cold water. Dissolve in a standard
potassium hydroxid solution 1 cc of which contains 10.405 mg of potassium hydroxid
and is equivalent to 0.25 mg of phosphorus. Titrate the excess with a nitric solution
1 cc of which is equivalent to 1 cc of the potassium hydroxid solution.
(c) Weigh 200 grams of the air-dried soil into a 2-liter flask, add 1,000 cc of distilled
water, and treat as in (5). Evaporate 750 cc of the clear solution to 50 cc and deter-
mine potassium as in fertilizers.^
(c/) Make a preliminary digestion of the air-dried soil in fifth-normal nitric acid at
room temperature. « AVeigh 220 grams of soil into a 2J-liter flask and add 2,200 cc of a
solution of nitric acid of such strength that after allowing for the nitric acid neutralized
by the soil, as shown by the previous digestion, there will be left 2,200 cc of fifth-
aU. S. Dept. Agr., Bureau of Soils, Bui. 31, p. 12.
&Ibid., p. 16.
cXJ. S. Dept. Agr., Bureau of Chemistry, Bui. 46, p. 11.
dibid, p. 22.
«Ibid., p. 74, (b).
143
normal nitric acid solution. Digest five hours at room temperature, shaking every
half hour. At the end of the digestion period shake and filter through a large folded
filter, pouring the solution back onto the filter until it runs through clear.
Evaporate 1,000 cc of the filtrate to dryness, moisten with hydrochloric acid, luring
to dryness again, bake five hours at 110°, take up witli hydrochloric acid and water,
filter, and wash. Evaporate filtrate and washings to about 25 cc, make alkaline with
ammonia, and determine phosphorus as in (6).
Evaporate two 500 cc portions of the filtrate to complete dryness and determine
total alkalies and potassium according to the official method^, beginning with "dis-
solve in about 25 cc of hot water, add an excess of baryta water." Complete one of the
two determinations according to this method, in the second introduce the modification
suggested in the Journal of the American Chemical Society.^
The directions given under (d) for the determination of phosphorus are not exactly
those followed by the Kentucky Experiment Station, some details being changed to
make the results comparable with the rest of the work on phosphorus.
The following laboratories have reported some work in time for this report:
The Bureau of Soils, reported by G. H. Failyer.
The Virginia Experiment Station, reported by W. B. Ellett.
The Arkansas Experiment Station, reported by J. H. Norton.
The Minnesota Experiment Station, reported by H. Snyder.
The Michigan Experiment Station, reported by A. J. Patten.
The Kentucky Experiment Station, reported by S. D. Averitt.
The Illinois Experiment Station, reported by J. H. Pettit.
Table 1. — Water-soluble phosphorus and potassium.
[Parts per million in dry soil.]
Analyst.
Soil No. 1.
Soil No. 2.
Sou No. 3.
Soil No. 4.
P.
K.
P.
K.
P.
K.
P.
K.
S. D. Averitt, Kentucky a
1.1
2.0
bO
1.7
6 12.1
/ 1.0
i::"
23.0
27.5
40.3
23.2
"i2."i"
12.5
0.9
2.0
6 5.0
.9
6 17.9
1.2
1.2
11.0
34.0
27.5
6 55.1
"io.'g'
10.4
0.9
2.0
6.1
2.7
.......
1.0
12.0
32.0
31.6
6 46.7
"is.G
12.8
2.3
3.0
6.4
2.3
6 21.2
2.5
2.5
G. H. Failyer, Bureau of Soils, Washington,
B.C.a. .....
23.0
42.5
J. A. Hummel, Minnesota a .
35.3
Doroethea Moxness, Michigan a
50.1
.T. H. Pettit, Illinois
A. Ystgard, Illinois '..
'"i4.'2
15. 1
Average
1.4
23.1
1.2
18.8
1.5
20.4
2.5
30.0
o Duplicates not reported.
6 Not included in average.
Considerable variation is to be noted among the results of the different analysts, in
some cases the supposedly poorer soils showing the larger amounts of phosphorus or
potassium.
G. H. Failyer, of the Bureau of Soils, determined phosphorus and potassium upon
these soils by the colorimetric as well as the gravimetric method. His results are
given in Table 2.
«U. S. Dept. Agr., Bureau of Chemistry, Bui. 46, p. 75 (k).
&1903, 25: 496.
144
Table 2. — Water-soluble phosphorus and potassium (Failyer).
[Parts per million in dry soU.]
Soil
No.-
Pota
ssiimi.
Phosphorus.
i Gravi-
metric
method.
Colori-
metric
method.
Gravi- Colori-
metric metric
method, method.
1
3
4
23
11
12
23
22
16
16
20
2 1
2 i 1
2 1
3 2
It Tvill be noted that the phosphorus results show some uniform differences Tvhile
those for potassium do not. It should be stated perhaps that phosphorus controls the
yield upon all of these soils in the field so far as the two elements under consideration
are concerned.
Table 3. — Phosphorus, soluble in jifth-normol nitric acid.
[Parts per million in <iry soU.]
Analyst.
Soil Soil Soil i Soil
No. 1. No. 2. No. 3. 1 No. 4.
S. D. Averitt. Kentucky.
W. B . Ellett. Virginia
J. H. Norton. Arkansas a
A. D. Wilhoit. Minnesota
J. A. Hummel. Minnesota i
Doroethea Moxness. Michigan a
J. H. Pettit. •Illinois
15.0
&22.0
18.0
1S.0
14.4
14.0
13.4
12.9
14.0
10.6
16.7
14.4
14.6
Average.
14.7
8.0
5.0
158.0
6 15.0
blQ.O
b 175. 0
10.0
0. I
158.0
n.o
6.0
158.0
7.4
4.4
165.5
7.4
4S
162.0
o.S
3.6
155. 5
6.1
2. S
154.5
7 7
3.1
163.4
5.7
3.6
145.6
5. 6
9.7
6.4
157.1
6 15.2
6 25.5
167.4
7.8
4.8
165.4
7. S
4.8
166.0
159.
a Duplicates not reported.
6 Not included in average.
[Note.— Aveiitt, No. 1. Work done according to directions sent out: No. 3. Baking of residue
omitted: No. 5. Baking omitted and phosphorus determined by Kentucky method.]
Table 3 shows in the main satisfactory agreement in the results reponed. The rela-
tively large amount of phorphorus given by this solvent in Xo. 4 is to be noted, and
also the fact that the relative amounts of this element in the other three soils is directly
opposite to their relative productive capacities.
It is to be regretted that so few reports were received vn the potassium work.
Table 4. — Potassium, soluble in fifth-normal nitric acid, by tuo methods.
[Parts per milUon in dry soU.I
Analyst.
Soil No.
Offi-
cial.
Modi-
fied.
Sou No. 2. I SoU No. 3.
?oU No. 4.
Offi- ' Modi- Offi-
cial, fied. , cial.
Modi-
fied.
Offi-
cial.
Modi-
fied.
W. B. EUett Virginia
J. n. Norton. Arkansas a
J. A. Hummel. Minnesota a
Doroethea Moxness, Michigan a.
A. Ystgard. Illinois
7 99.2
\ 88.7
113.0
95.5
89.3
f 91.3
t 96.1
95.9
95.0
130.0 i 142.9 ! 141.
135. 9 I 139. 8 f 139. 8
128.0 i 6188.0
132. 9 236. 0
138. 8 233. 1
6 276.0
135 5 ' 234. 4
679.0 !6107.§ 697.0 !6mS "123.4' 6 18911
97. 7 136. 0 139. 7 ' 142. 5 i 150. 4 244. 9
94. 5 133. S 137. 1 139. 8 145. 9 249. 9
Average .
230.7
235.1
6 1S7.0
253.2
250.5
96. 2 95. S 131. 9 139. 9 139. 9 j 142. 0 ) 2-39.
o Duplicates not reported.
6 Not included in average.
145
It is to be observed that in all cases the amount found is in proportion to the pro-
ductiveness of these soils. It is to be further noted that in two of the three cases
where results by both methods are reported the two methods give concordant results.
Comments of Analysts.
Andrew J. Patten: The water solutions were filtered by means of suction through a
thick pad of asbestos in a Buchner filter. The acid solutions were filtered through
paper. The low results for total phosphorus by fusion with sodium peroxid is probably
due to an incomplete fusion, as in her desire to follow the method strictly, Miss Mox-
ness was very careful not to bring the mixture to complete fusion, and this is thought
to be the reason for the low result. She did not have time to duplicate this work and
so verify this point.
S. D. Averitt: In regard to your instructions for the determination of phosphorus,
I beg to offer two objections: (1) Our experience in this laboratory indicates that
the prolonged baking before filtering from the silica is not only unnecessary and a loss
of time but that it renders the subsequent solution and filtration more difficult; (2)
after the addition of molybdic solution and keeping at 50° to 60° for two hours, the
standing over night seems open to serious objection, on account of the probable pre-
cipitation of molybdic acid, especially if a large excess of molybdic solution has been
added. In the first determinations under (d) aliquots corresponding to 100 grams of
soil were taken, and in the second determination aliquots corresponding to 40 grams of
soil only could be had, but the same quantity' of molybdic solution was used in each
case. Now, in Nos. 1, 2, and 3, in which the amount of phosphorus is small, the second
determinations show a large increase over the first; in No. 4, in which the amount of
phosphorus present is from ten to thirty times as great as in 1, 2, and 3, the increase
relatively is not nearly so great.
Harry Snyder: I note that in the case of the soils that contain appreciably large amounts
of phosphorus soluble in fifth-normal nitric acid reasonably concordant results are
secured, and in the case of soils with small amounts the agreement between the work
of the two analysts is approximate. The difference is, I think, due to the fact that the
end point in the titration is not as perfect as could be desired, and one analyst possibly
carries the titration further than the other. In the case of large amounts of phos-
phorus this does not appear to affect the results appreciably as in the presence of
small amounts.
Methods for the Determination of Total Phosphorus.
Directions for determining the total phosphorus by a sodium peroxid method to be
compared with the usual alkali fusion method were also sent out. According to these
directions, for the sake of greater accuracy, it was suggested that a larger amount of
material be used than that suggested in our paper on the method presented last year.
Later it was found that the changes made necessary by this modification affected the
method fundamentally, giving much lower results and making the method cumber-
some. Accordingly these directions and the results obtained thereby are omitted
from this report.
The four samples were run for total phosphorus in the Illinois lalxjratory, however,
by the following method :
Weigh 10 grams of sodium peroxid into an iron or porcelain crucible and thoroughly
mix with it 5 grams of the soil. If the soil is very low in organic matter, add a little
starch to hasten the action. Heat the mixture carefully l)y applying the flame of a
Bunseh burner directly upon the surface of the charge and the sides of the crucible until
the action starts. Cover crucil)le until reaction is over and keep at a low red heat for
fifteen minutes. Do not allow fusion to take place. By means of a large funnel and
a stream of hot water, transfer the charge to a 500 cc measuring flask. Acidify with
hydrochloric acid and boil. Let cool and make up to the mark. If the action has
31104— No. 105—07 10
146
taken place properly there should be no particles of undecomposed soil in the bottom
of the flask. Allow the silica to settle and draw off 200 cc of the clear solution.
Precipitate the iron, alumina, and phosphorus with ammonium hydroxid; filter,
wash several times with hot water, return the precipitate to the beaker with a stream
of hot water, holding the funnel over the beaker, and dissolve the precipitate in hot
hydrochloric acid, pouring the acid upon the filter to dissolve any precipitate remain-
ing. Evaporate the solution and washings to complete dryness on a water bath.
Take up with dilute hydrochloric acid, heating if necessary, and filter out the silica.
Evaporate filtrate and washings to about 10 cc, add 2 cc of concentrated nitric acid,
and just neutralize with ammonium hydroxid. Clear up with nitric acid, avoiding
an excess. Heat to 40° to 50° on water bath, add 15 cc of molybdic solution, keeping at
this temperature one to two hours. Let stand over night, filter, and wash free of
acid with a one-tenth per cent solution of ammonium nitrate and, finally, once or
twice with cold water. Transfer filter to beaker, and dissolve in standard potassium
hydroxid (1 cc=0.2 mg P), titrate the excess of potassium hydroxid with standard
nitric acid, using phenolphthalein as indicator.
These details differ but slightly from those presented a year ago. The method was
checked by the alkali-carbonate method, and Table 5 gives the results obtained.
Table 5. — Total phosphorus sodium-peroxid and
nois station).
dkali-carhonate fusion methods {Illi-
[Parts per million in dry soil.]
Soil
No.
Sodium-
peroxid
fusion.
Alkali-car-
bonate
fusion.
1
2
3
4
487.8
437.1
477.6
859.3
■
492.9
432.0
469.4
, 846.9
In November, 1905, practically the same directions were sent to a number of the
members of this association who had consented to test the method. With these were
sent two samples, one of soil and the other of the same soil to which rock phosphate had
been added, so that the percentage of phosphorus in the mixture was 0.080. Table 6
contains the results of this work.
Table 6. — Total phosphorus, sodium-peroxid fusion method (cooperative ivork).
[Per cent in air-dried soil.]
Analyst.
Soil.
Soil plus
rock
phos-
phate.
A. Ystgard, Illinois . .
0.037
.036
.046
.046
.038
.038
0.081
R. Harcourt Canada
.060
A . M. Peter, Kentucky
.086
B. W. Kilgore North Carolina
.081
.082
J. W. Ames, Ohio
.071
From the results obtained this year and from others which have come under his obser-
vation, the referee does not recommend further work on water-soluble plant food, but
would suggest rather that methods for the rapid determination of the total amount of
the plant-food elements present be investigated, in order that we may thereby deter-
mine systems of extensive farming which may be applied over longer periods. In
addition to this, further study should be made of some weaker solvent — fifth-normal
nitric acid seems to be desirable — in order that we may thereby obtain some idea of
the amount of the plant-food elements which may be obtained in shorter periods, and
thus of the treatment necessary for intensive cropping of the land. Accordingly your
referee would make the following recommendations:
147
Recommendations.
It is recommended —
(1) That the fifth-normal nitric-acid digestion method be further studied, both by
correcting for the basicity as shown by a previous digestion and without such correction.
(2) That the sodium-peroxid fusion method for total phosphorus be given further
trial and that this be compared with the alkali-carbonate fusion method.
(3) That the modified J. L. Smith method for total potassium presented at this meet-
ing be further tested.
(4) That line 30 under "1. Preparation of sample," page 71, Bulletin No. 46, be
changed from "openings one-half millimeter in diameter" to "openings one millimeter
in diameter," and that "passed through a sieve of one millimeter mesh" be omitted
from line 1 under (h) page 74, Bulletin No. 46.
(5) That the "Determination of volatile matter," page 72, Bulletin No. 46, be
replaced by the " Determination of total organic carbon." (J. Amer. Chem. Soc, 1904,
26 : 1640.)
(6) That the "Determination of manganese," page 73, Bulletin No. 46, be omitted.
(7) That under (k), page 75, Bulletin No. 46, mark the official method "(a)" and
insert the following:
B. OPTIONAL PROVISIONAL METHOD.
Proceed as in (a) through "let stand a few minutes in the water bath" and complete
as follows:
Filter into a beaker, add a drop or two of hydrochloric acid and 1 cc of ammonium
sulphate (75 grams to 1 liter), digest several hours on water bath, and filter into a tared
platinum dish. Evaporate to complete dryness, heat to full redness, add 1 gram of pow-
dered ammonium carbonate, expel by heating, cool, and weigh the sulphates of sodium
and potassium. Determine potassium in the usual manner.
DETEEMINATION OF TOTAL POTASSIUM IN SOILS.
By J. H. Pettit and A. Ystgard.
The following method is to be used for total potassium only and not for total alkalis.
The well-known ammonium-chlorid and calcium-carbonate fusion, devised by J. Law-
rence Smith, o is used. The fused mass is transferred to a porcelain dish, slaked with
hot water, finely ground with an agate pestle, and transferred to a filter. After wash-
ing free of chlorids, the filtrate and washings are concentrated in a Jena beaker to about
20 cc and filtered. Filtrate and washings are slightly acidified with hydrochloric acid,
concentrated in a platinum dish, and l^ cc of a platinic chlorid solution (10 cc contains 1
gram of platinum) added where 1 gram of soil has been used. This is then evaporated
to a sirupy consistency as usual and washed with 80 per cent alcohol and ammonium
chlorid solution.
In Table 1 are given the results obtained upon five soils, first, when all calcium was
removed as in the regular J. L. Smith method; second, when part of the calcium was
removed by one precipitation with ammonium carbonate; and third, when no cal-
cium was removed except that thrown out upon evaporating as indicated in the
method described above.
« Fresenius's Quantitative Analysis, p. 426.
148
Table 1. — Total potassium determined by the Smith method and modification.
Serial
number.
AU calcium
removed
(Smith).
Bulk of
calcium re-
moved by
one pre-
cipitation.
No calcium
removed.
1 Per cent.
1263 1.731
1264 1 1.794
1265 1.843
1266 1.853
1267 1 2.064
1
Per cent.
1.716
• 1. 781
1.818
1.847
2.038
Per cent.
1.736
1.785
1.840
1.848
2.037
With the Smith method without the removal of lime one man can easily make ten
determinations on the average in a day of eight hours. Table 2 contains the results in
duplicate obtained upon a number of soils.
Table 2. — Potassium determinations in duplicate by the J. L. Smith method as modified
by the authors {without removal of lime).
Soil No.
Potassium (K).
Soil No.
Potassium (K).
Per cent.
Per cent.
Per cent.
Per cent.
480
1.235
1.240
780
1.543
1.549
481
1.295
1.290
781
1.488
1.497
482
L232
1.248
782
1.563
1.563
483
1.296
L296
783
1.594
1.633
484
1.362
1.385
784
1.578
1.589
485
1.353
1.353
785
L560
1.560
EEPOET of committee on EEVISION or METHODS.
The President called for the report of the committee on the revi-
sion of methods and the secretary asked permission to explain the
appointment and purpose of the committee before the report was
submitted.
Mr. Wiley. The association at the last meeting urged the secretary to issue a revised
edition of Bulletin No. 46 on methods of analysis, instructing him to insert such
changes as had been authorized by the association since the issuance of the methods
in 1899 and make such verbal changes as were necessary to insure uniformity. With
this end in view a study was made of the methods and the additions thereto by a com-
mittee composed of the chiefs of laboratories of the Bureau of Chemistry, from which
it appeared that the revision as contemplated was not practicable for the following
reasons:
The growth in the work, and the far-reaching changes and additions made since
the issuance of the methods of analysis in 1899 and of the provisional methods of food
analysis as Bulletin No. 65 in 1902, and the overlapping of these methods in many
instances made it impossible to prepare a satisfactory revision without entirely rear-
ranging the methods and, consolidating these two bulletins. It was deemed wise
therefore in the interests both of efficiency and of economy not to publish a partial
revision prior to this convention but to ask the president to appoint a committee of
the association to make a complete revision and submit it to the association for approval.
This has been done, and the committee which has been at work previous to this meet-
ing is now ready to report.
In partial compliance with the instructions of the association Circulars 28, 29, and
30 have been issued from the Bureau of Chemistry, .giving the authorized revision
149
of the provisional methods for the determination of food preservatives, and the changes
in and additions to Bulletins 46 and 65, respectively, since those reports were issued.
Mr. Haywood, chairman of the committee on the revision of methods
submitted the following report, together with the manuscript of the
revised methods:
The committee recommends that their report as submitted be provisionally accepted
by the association for final adopton at the next meeting and that provision be made to
supply the members of the association with copies for the purpose of criticism, before
the next meeting, which criticism shall be submitted to the secretary in time for suffi-
cient consideration before the meeting in 1907.
A number of minor recommendations submitted for the guidance
of the secretary in editing the revision and insuring uniformity of
expression were also submitted by the committee. The more impor-
tant of the recommendations are as follows:
Recommendations of Revision Committee.
It is recommended that —
(1) A table of atomic weights be inserted.
(2) The secretary shall see that the methods are definitely designated either by the
name of the originator or by some prominent reagent used in connection with them,
especially where there are a number of methods for the same determination.
(3) All the methods be published in one bulletin.
(4) All weights and measures be expressed according to the metric system.
(5) The word ''water" whenever used without qualification means distilled water;
call attention to this in some prominent place in the bulletin.
(6) The strength of strong alcohol be expressed uniformly either as 95 or 96 per cent;
dilute alcohols to be expressed both in percentage strength and specific gravity.
(7) Krober's table for pentosans be inserted.
(8) Describe Volhard's method for chlorin under ash.
(9) Certain methods for insecticides be made provisional by the association.
(10) The Ruffle and soda-lime methods be dropped as official methods.
(11) The heading "Infants' and invalids' foods" be dropped.
(12) Under "Distilled liquors, Detection of methyl alcohol," the removal of acetal-
dehyde by Prescott's method be dropped; under same heading drop the S. P. Mulliken
method for the removal of formaldehyde.
(13) The methods under "Dairy products" for the determination of casein monolac-
tate and casein dilactate be dropped.
(14) Under "Flavoring extracts" the phenylhydrazin hydrochlorid method be
dropped.
(15) The first method for the approximate quantitative estimation of saccharin be
dropped.
(16) Methods for cocoa and chocolate be inserted as provisional.
(Signed) J. K. Haywood, Chairman.
J. P. Street.
F. W. WOLL.
J. H. Pettit.
L. M. Tolman.
F. P. Veitch.
A. L. Winton.
150
On motion b}" Mr. Patrick the provisional method kno^^^l as the
Water house test, for the detection of oleomargarin and renovated
butter, was also dropped from the revised methods.
]Mr. Wiley, referring to the main recommendation in regard to the
distribution of the revision of the methods for criticism, called atten-
tion to the fact that so extensive a piece of printing could hardly be
done for temporary use only, and asked for an opmion in regard to
the changes that would be necessary before adoption. Mr. Hay-
wood replied that as far as the committee was concerned their action
was final and it was hoped that onlj^ minor corrections would be made.
With this understanding the report of the committee was adopted
by the association, and upon motion by Mr. Bigelow the committee
on revision of methods was continued for anotjier 3^ear.
EEPOET or COMMITTEE Olf NOMINATIONS.
Mr. Woods, chairman of the committee on nominations, presented
the following report :
For president, Mr. J. P. Street, of New Brunswick, N. J. ; for vice-
president, Mr. Harry SuA^der, of St. Anthony Park, Minn. ; for secre-
tary, Mr. H. W. Wiley, of Washington, D. C. ; as additional members of
the executive committee, Mr. B. B. Ross, of Auburn, Ala., and Mr.
B. L. Hartwell, of Kingston, R.I.
The secretary was instructed to cast a unanimous vote for the
officers nominated.
THIRD DAY.
FRIDAY— MORNING SESSION.
EEPOET ON INORaANIO PLANT CONSTITUENTS.
By W. W. Skinner, Referee.
The work for this year has been confined to an investigation of sulphur and phos-
phorus found in plant materials, and the amounts of these substances remaining in
the ash as obtained by various 'methods. The two materials selected for the work
were cotton-seed meal (sample A), and white mustard-seed meal (sample B).
Total sulphur was determined by the peroxid method, and also by the absolute or
combustion method (proposed by Sauer« and modified by Tollens and Barlow &)
the results by the two methods agreeing fairly well. The results by the peroxid
method are so entirely satisfactory, and the method has been given such a thorough
test, that the referee recommends that this method for total sulphur, which is now
provisional, be adopted by the association as an official method.
The results by the combustion method are interesting and it is believed that further
investigation of the volatile and nonvolatile products obtained by this method may
throw considerable light upon the vexing problem of how best to consider those
so-called inorganic constituents which are partially or wholly lost in the ordinary
methods of ashing plant substances. It is therefore recommended that particular
attention be devoted by the referee for next year to this line of investigation.
Two years ago the association voted to substitute for "ash" the title of "inor-
ganic plant constituents, " which to the referee seems unfortunate as it leads to much
confusion of terms. In our methods for the proximate analysis of foods and feeding
stuffs, we have a distinct separation known as ash, yet nominally we havoNUO methods
for its analysis, unless "ash" is to be considered synonymous with "inorganic plant
constituents." To the referee it seems that there is ample room for work upon ash,
meaning thereby the residue obtained^ by ignition of organic materials in air or in
oxygen, under approved and standard conditions, in addition to the investigation of
the so-called inorganic plant constituents, some of which are volatilized, in part at
least, by one of the above methods of combustion.
The following tables and discussion show the nature of the work for the year. Five
chemists signified their intention of collaborating in the work and samples were sent
to each in March, but no reports upon the samples have been received. The results
are therefore confined to those obtained by the referee and one collaborator in the
Bureau of Chemistry.
aZts. anal. Chem., 1873, p. 32.
& J. Amer. Chem. Soc, 1904, 26: 341.
(151)
152
Table 1. — Ash, sulphur (SO^), and phosphorus (PO^) in moisture-free cotton-seed meal
(A) and mustard-seed meal (5\
Method.
Analyst.
Ash.
Sulphur (SOi).
Phosphorus
(PO4).
A.
B.
A.
B.
A.
B.
' Per ct.
W. W. Skinner 1
Per ct.
Per ct.
j 0.17
t .13
Per ct.
0.33
.30
Per ct.
""4.48
Per ct.
2.01
C. Goodrich
/
( 7. 54
4.35
4.33
2.04
C.Goodrich •^ - ._
Ashing with calcium acetate.
W. W. Skinner
C. Goodrich
C. Goodrich
W.W. Skinner
j .37
\ .38
1.91
1.84
4.56
4.51
2.07
r 7.76
\ 7.74
7.67
I 7.61
1 7.53
I 7.55
5.67
5.63
4.19
4.36
4.19
4.29
2.07-
Combustion method
1.64
1.61
4.24
4.06
4.36
4.55
4.38
2.04
1.87
Peroxid method
W. W. Skinner
/ 1.73
\ 1.74
j 1.69
1 1.67
4.37
4.43
4.33
4.34
4.49
4.54
4.53
4.56
1.96
C. Goodrich
1.94
2.01
1.93
Table 2. — Volatile and nonvolatile sulphur (SO^) determined by the combustion method
on moisture-free matej'ial.
Sample.
Volatile
sulphur.
Nonvola-
tile sul-
phur.
Total
sulphur.
A
B
Per cent.
f 1.62
1 1.56
4.11
\ 3.79
[ 4.15
Per cent.
0.02
.05
.13
27
!21
Per cent.
1.64
1.61
4.24
4.06
4.36
Discussion of Results.
Ash. in the two samples was determined by tln^ee methods:
(1) The ordinary method by charring, extracting with water, igniting the residue,
adding the extracted matter, evaporating to drjTiess, etc.
(2) By igniting with a known quantity of calcium acetate.
(3) By combustion in oxygen.
A very close agreement is noted in the results for ash in samples bm'ned in air and in
oxygen. The sulphur in the ash obtained by these two methods is also fairly compar-
able. The volatile sulphur determined by the combustion method added to the sul-
phur remaining in the ash. obtained by this method gives for total sulphur, figiues
comparable with total sulphur obtained by the peroxid method; the averages on sam-
ples A and B, respectively, are 1.49 per cent against 1.56 per cent, and 3.97 per cent
against 4.11 per cent. Comparing the sulphur in the ash obtained by the usual method
of ashing with the total sulphur obtained by the peroxid method a loss is noted of 91.2
per cent in the case of sample A and 92.7 per cent in the case of sample B. Practically
all of the phosphorus remains in the ash, only traces being volatilized when proper
precautions are exercised to prevent too high a temperature and fusion of the ash.
Recommendations.
It is recommended:
(1) That the peroxid method for total sulphur be adopted as official.
Peroxid method. — Place from 1.5 to 2.5 grams of material into a nickel crucible of
about 100 cc capacity and moisten with approximately 2 cc of water. Mix thoroughly,
using a nickel or platinum rod. Add 5 grams of pure anhydrous sodium carbonate and
mix. Add pui-e sodium peroxid, small amounts (approximately 0.50 gram) at a time,
153
thoroughly mixing the charge after each addition. Continue adding the peroxid until
the mixture becomes nearly dry and quite granular, requiring usually about 5 grams of
peroxid. Place the crucible over a low alcohol flame (or other flame free from sulphur)
and carefully heat, with occasional stirring, until contents are fused. (Should the
material ignite the determination is worthless.) After fusion remove the crucible,
allow to cool somewhat, and cover the hardened mass with peroxid to a depth of about
0.5 cm. Heat gradually, and, finally, with full flame until complete fusion, rotating
the cnicible from time to time in order to bring any particles adhering to the sides into
contact with the oxidizing material. Allow to remain over the lamp for 10 minutes
after fusion is complete. Cool somewhat. Place warm crucible and contents in a
600-cc beaker and carefully add about 100 cc of water. After violent action has ceased
wash material out of crucible, make slightly acid with hydrochloric acid (adding small
portions at a time), transfer to a 500-cc flask, cool, and make to volume. Filter and take
a 200-cc aliquot for determination of sulphates by precipitating with barium chlorid in
the usual manner.
(2) That the combustion method (i. e., the Sauer-Tollens-Barlow method) for deter-
mining volatile inorganic plant constituents be further investigated.
Mr. Wiley. It seems to me that the referee on inorganic plant con-
stituents should not only study the proposed methods of analysis but
should also consider the question, What is ash? The question arises,
Is sulphur in combination an inorganic constituent? If it is, then
nitrogen in inorganic compounds is an inorganic constituent also, and
the same is true of every one of the elementary substances. The pro-
priety of determining total sulphur in plants no one questions, but is it
an inorganic constituent as determined by these methods? In my
opinion, ash is the residual matter which remains after ordinary com-
bustion of substances in the atmosphere, conducted in such a manner
as to avoid loss by volatilization of the real ash constituents, and to
secure under proper forms as complete a combustion of the carbon as
possible. When incineration is conducted in that way a very large
part, sometimes all, of the organic sulphur escapes detection and the
same is true, to a large extent, of phosphorus. Therefore it appears
that only the organic sulphur and phosphorus that can be fixed with
the natural bases present before or during combustion can claim any
place in the ash. It is very important in these studies that a clear dis-
tinction be made between the term '^ ash' ' as generally defined and the
total sulphur, phosphorus, or other substances which may be present
in inorganic form. The issue has become acute in some cases as to
whether or not a determination of: ash by the ordinary method really
determines all of the ash. Some analysts are of the opinion that the
ash should include the total sulphur and phosphorus in any form, and
unless that is effected a real ash determination is not made. I do not
share that opinion, simply from the analytical point of view, but since
these difi'erences of opinion exist it seems advisable that the associa-
tion instruct the referee for the coming year to take up the question,
^'Whatisash?"
154:
EEPOET OF COMMITTEE B ON EEOOMMENDATIONS OP KEPEEEES.
By E. B. HoLLAXD. Chairman.
(1) Medicixal Plants and Drugs.
It is recommended:
1. That the present provisional method for the assay of opium products be made
official. [Proceedings, 1905, Bui. 99, p. 162; U. S. Pharmacopa?ia, eighth revision,
p. 329, with additions.]
RefeiTed to the referee for 1907 for recommendation and final action by the
association.
2. That the present line of work be continued by the referee on medicinal plants and
drugs.
Adopted.
(2) Foods and Feeding Stuffs.
It is recommended:
1. That the Konig and modified Konig method for crude fiber be dropped. [See
Proceedings, 1903, Bui. No. 81, p. 38.]
Adopted.
2. That the slightly modified Ellett-Tollens method for methyl-pentosans be tested
again next year by the association. [J. Landw., 1905, 53, (1): 13.]
Adopted.
3. Upon motion by Mr. Peter the association voted to instruct the referee for 1907 to
study the temperature and time of drying of ether extracts.
Passed.
(3) Dairy Products.
It is recommended : .
1. That the method of double extraction be adopted as provisional for the determina-
tion of fat in condensed milk, the directions to be given as follows:
Extract the solid residue of about 5 grams of a 40-per cent solution with ether in the
usual manner; dry, leave tubes in a dish containing 500 cc of water or more, for two or
three hours; drv, extract again for about five hours, and determine fat as under milk.
[Bui. 46, p. 54^]
Adopted.
2. That the Gottlieb method [Landw. Versuchs. Stat., 1892, 40: 1] be made pro-
visional for the determination of fat in milk.
RefeiTed to referee for 1907 for further recommendation.
3. That the conversion factor for protein in milk and dairy products be changed to
6.38 throughout the methods.
Referred to committee on unification of terms for reporting anahi:ical results.
4. That the study of methods of analysis of condensed milk be continued with
special reference to the determination of lactose and sucrose in the sweetened product.
Adopted.
(4) Sugar.
recommendations pending from 1905.
The following recommendations are made in regard to questions referred from 1905
for action [see Proceedings, 1905, Bui. 99, pp. 18-25, p. 155; Cu'cular 26, p. 5J:
155
Chemical methods.
1. That the methods selected by the international committee for unifying sugar
analysis, now provisional, be made official with the exception of paragraph 7, which
should be eliminated, and paragraph 8, upon which action should be deferred. [See
Proceedings, 1902, Bureau of Chemistry Bulletin No. 73, p. 58.]
Adopted.
2. That action upon recommendations 3 and 4 be deferred until 1907. These
recommendations are as follows:
3. That method (a) for the determination of copper in the cuprous oxid precipitate,
requiring reduction in hydrogen [see Bulletin No. 46, p. 37], be dropped as an official
method of this association.
4. That methods (c) and (d) for the determination of copper in the cuprous oxid
precipitate, requiring the electric deposition of the metal [see Bulletin No. 46, p. 38],
be dropped as official methods of the association.
Action deferred as recommended.
3. That action for the adoption of Low's thiosulphate method for determining copper
in the cuprous oxid be deferred until the original method and its modification has been
compared by the association.
Adopted.
Molasses methods.
Of the five recommendations deferred from 1905 for action, No. 1 is covered by
recommendation No. 3 under general recommendations on sugar, 1906, which follows,
while action on Nos. 2, 3, 4, and 5 is deferred until further cooperation is obtained
from interested chemists. [See Circular 26, p. 5, or Bulletin No. 99, p. 24.
GENERAL RECOMMENDATIONS, 1906.
It is recommended:
1. That further work be done upon the comparison of moisture methods for molasses
and massecuites before action be taken by the association.
Adop.ted.
2. That when increased accuracy is desired in the determination of reducing sugars
in commercial products on account of the danger of contamination in the precipitated
cuprous oxid, only such methods be followed as provide for a direct determination
(electrolytic or volumetric) of the reduced copper.
Adopted.
3. That in the optical methods for the analysis of dark-colored massecuites, molasses,
and other low-grade products, the normal weight of substance may be made to some
multiple of 100 cc, according to the volume necessary for accurate polarization. This
normal weight may be weighed out directly and made up to the desired volume, or
taken from a solution prepared by dissolving a larger amount of the material (50 to
100 grams) with water to a definite weight or volume. The latter method is to be
preferred with nonhomogeneous materials, such as massecuite, to secure greater uni~
formity of sample.
Adopted.
4. That in the estimation of sucrose by the gravimetric method the calculation by
the factor 0.95 (Bulletin 46, p. 39, bottom line) be made only upon the reducing
sugars estimated as invert sugar. In making this calculation, reducing sugars pre-
viously present which require to be subtracted must also be expressed as invert
sugar.
Adopted.
5. Provisional: That reducing sugars in mixtures either before or after inversion
may also be expressed as dextrose, in which case the calculation of the true percent-
ages of sucrose and other sugars may be made by appropriate conversion factors and
formulae. [Browne, Analysis of Sugar Mixtures: J. Amer. Chem. Soc, April, 1906,]
156
6. Provisional: That for the estimation of dextrose, invert sugar, maltose, and lac-
tose, under uniform conditions of analysis, the method and table of Munson and ^Valker
be adopted provisionally by the association. [Alunson and Walker. Unification of
Method for Reducing Sugars: J. Amer. Chem. Soc, June. 1906.]
Upon the suggestion of the referee, action on recommendations 5 and 6 was deferred,
since but little cooperative work had been done.
7. That the questions of the influence of the lead precipitate, and the use of hypo-
chlorites, hydrosulphites, and other agents for decolorizing in optical analysis, and
the use of clarif\'ing agents before estimating redticing sugars, be iuA'estigated further
before decisive action is taken by the association.
Adopted.
8. That the cooperation of sugar chemists not members of the association be invited.
to secure a greater expression of opinion upon matters pertaining to improvement of
anal}i;ical methods.
Adopted.
(5) Taxxix.
1. The report of the referee on tannin was not received until after the report of
Committee B had been made, but it was considered by the committee and a supple-
mental report made by Mr. Kebler in regard to the recommendation of the referee
that "the tannin section of the Association of Official Agiicultural Chemists abandon
the appointment of referee and associate referee and omit supervision of the tannin
method." The committee recommended that the tannin work of the association be
continued, and the recommendation was adopted by the association.
2. Upon motion by Mr. Veitch the method on tannin submitted by the referee in
1905 [Bui. 99. p. 123]. as revised by the committee on revision of methods, 1906, was
adopted as a provisional method, the changes made by the committee being almost
exclusively in manner of statement, etc.
Ill discussing the sugar recommendations ^Ir. Wiley made the fol-
lowmg remarlvs: The reconimendation that all polarizations be made
at 20° C. must be interpreted to mean that when it is not practica-
ble to maintain tliis temperature the results should be referred to it
for correction. Tliis is necessary because in commercial work the
wide differences in temperature wliich occur result m enormous dif-
ferences in polarization. As I showed some years ago, a difference of
1.25 per cent may be obtained on the same sugar by polarizing at
different temperatures. The members of the association may recall
that a legal battle was fought on tliis point on the ground that when
Congress passed a law sa^-ing ^^as determined by polarization'' the
system of polarization then ui commercial use was meant, in which
no attention was paid to the temperature of polarization. The ques-
tion was carried to the Supreme Court of the United States and the
decision of the lower court sustained, namely, that the Secretary of
the Treasmy had absolute right to prescribe the regulations under
which the polarization should be made, and to say that it must be
made under the conditions obtaining when the law was passed was
to bar scientific progress. Thus the principle was established that
these corrections for temperature could be legally applied in commer-
cial work.
157
Mr. WoLL. I would like to bring up a matter of general interest.
It seems only right that the association should take action in regard
to the views expressed in the president's address, and to that end,
Mr. President, I move that a committee of seven on the presidential
address be appointed by the vice-president, to consider the questions
discussed in the address and to make such recommendations as to an
expression of the position of this association as in their judgment are
deemed advisable.
The motion was carried.
EEPOET 0^ PHOSPHOEIG ACID.
By B. W. KiLGORE, Referee.
Three questions were submitted by the association to the referees on phosphoric acid
for investigation the past year, as follows:
(1) Methods for determining available phosphoric acid in Thomas or basic slag.
(2) Methods for determining iron and alumina in phosphates.
(3) Testing citrate solution.
Mr. McCandless, the associate referee, agreed to undertake the investigation of prob-
lems (2) and (3), leaving for our consideration methods for determining available phos-
phoric acid in slag. For four or five years we have been conducting field experiments
with slag in comparison with other phosphates on three types of soil — red clay, sandy
loam with red clay subsoil, and a sandy loam of the coastal plain section. The data
are being tabulated, but are not ready for publication.
We also have samples of the slag which we propose to analyze by different methods
for the purpose of comparing laboratory and field results. It is not possible, because
of the incompleteness of the work, for us to submit a report with recommendations at
this time, but it is our purpose to continue the investigation and report at a later date.
EEPOET ON DETEEMINATION OF lEON AND ALUMINA IN PHOSPHATES
AND ON THE OITEATE SOLUTION.
By J. M. McCandless, Associate Referee.
The unusual amount of official work imposed upon the associate referee during the
fertilizer season in Georgia made it impossible for him to send out instructions and sam-
ples for cooperative work, or to file a timely report with the secretary. Some work,
however, has been done which promises well and may serve as a basis for careful study
and comparative tests during the coming year. The two subjects which I proposed to
investigate as associate referee on phosphoric acid were the methods for the determi-
nation of iron and alumina in phosphate rock, and the neutralization of the ammonium
citrate solution. The following notes are offered on these subjects:
Mr. E. G. Williams, of the Georgia State laboratory, collaborated in the work on iron
and alumina. We made a number of analyses of phosphate rock by various methods,
chiefly the acetate and the Glaser methods, with their various modifications, with
varying results, so that when the work was finished we were not sure of the actual per-
centages of iron and alumina in any of the samples. Evidently it was necessary to
work on a solution containing known quantities and resembling a solution of rock phos-
phate as closely as possible. To that end a solution in hydrochloric acid was prepared,
containing known quantities of chemically pure calci\im carbonate, microcosmic salt,
ammonia-alum, and iron wire.
158
From this solution the best results were obtained by the acetate method, as modified
and described by Gladding, and by the original Glaser method. Both methods, how-
ever, involve the difficult washing of a gelatinous precipitate, and in both methods if
a separation is made the alimiina must be obtained by difference. From a considera-
tion of the combining weights of the phosphates of iron and alumina, it was evident that
the present commercial practice of dividing the weight of the mixed phosphates by
two to obtain the percentage of oxids of iron and aluminum might be the cause of gi*ave
errors, and that any really scientific method of making the determination must involve
the separation of the two metals in a state of purity.
In the endeaA^or to avoid the washing of the gelatinous precipitate an indirect method
suggested itself. An aliquot of the phosphate solution in which the phosphoric acid
had been carefully determined, was transferred to a 250 cc fiask. precipitated by the
the acetate method, and the solution made up to the mark, cooled, and filtered through
a dry filter. An aliquot of this solution was then boiled down with excess of nitric
acid until acetic acid was driven off, and the phosphoric acid carefully determined
in the residue; the difference between the phosphoric acid in the original solution
and in the acetate solution gave the phosphoric acid which had combined with the
iron and aluminum. Some promising results were obtained in this way, but the
method requires the utmost accuracy in the determination of the phosphoric acid.
The iron being determmed by a volumetric method, the alumina was of course easily
calculated. The results, however, were not sufficiently uniform and the following
method was finally decided upon, and is recommended, not only for its accuracy but
for its rapidity and ease of manipulation. It is based upon Wohler's observation,
that aluminum phosphate maybe separated from iron by means of sodium hyposulphite
in a solution, slightly acid with hydrochloric and acetic acids. The details of the
method are as follows :
Weigh out 2.5 grams of phosphate rock, dissolve in 25 cc of hydrochloric acid, make
up to 250 cc. and filter off 100 cc through a dry filter, neutralize with ammonia, clear
with hydrochloric acid, add 200 cc of water. 2 cc of concentrated hydrochloric acid,
10 grams of sodium thiosulphate. and 15 cc of acetic acid (specific gravity 1.04). Boil
for fifteen minutes, filter on an ashless filter paper, and wash with ammonium nitrate
sohition. Ignite over a low flame tmtil the filter paper is ashed and finally at a red
heat. Weigh as aluminum phosphate containing 41.85 per cent of alumininn oxid.
Determine iron m another portion of the original solution by the bichromate or other
suitable volumetric method.
The following solutions were prepared: Xo, 1 to represent a rock containing 2.5 per
cent of ferric oxid and 2.5 per cent of aluminum oxid; Xo. 2. 2 per cent of ferric oxid
and 5 per cent of aluminum oxid; Xo. 3, 5 per cent of ferric oxid and 2 per cent of
alumintim oxid; Xo. 4, 5 per cent of ferric oxid and 7 per cent of alumintim oxid. All
of the solutions were prepared to represent a rock containing 35 per cent of calcium
-oxid and 31 per cent of phosphoric acid.
The results on alumina were 2.62 per cent found, r. 2.5 per cent present in Xo. 1;
in Xo. 2, 5.10 v. 5; in Xo. 3, 2.15 v. 2; in Xo. 4, 7.14 r. 7. The precipitates agglome-
rated with the free sulphur on boiling, washed with ease, and the ignited residties
were perfectly white and free from reddish color. The results obtained volumet-
rically on the iron also agreed closely with theory. In the case of sample Xo. 2 the
ustial method of diAdding the weight of the mixed oxids, as obtained by the ordinaiy
acetate method or by the Glaser method, by 2 would involve an error of nearly
1 per cent.
In the investigation as to the neutralization of the official solution of ammonium
citrate my collaborator was Ish. Joseph Q. Burton, first assistant State chemist of Georgia.
Chemists have long been dissatisfied with both the litmus and corallin tests for
neutrality of the solution. The calcium-chlorid method only enables one to tell when
a neutral point is reached at the expense of repeated tests and much trouble. The
following method for making a neutral solution is based upon the fact that citric acid
159
may be titrated with accuracy by a standard solution of caustic soda or potash with
phenolphthalein as indicator.
First determine the percentage of pure citric acid in the commercial article with
standard caustic potash and phenolphthalein. Weigh out a suitable quantity, calcu-
late the quantity of ammonia solution necessary to neutralize it, pour the somewhat
diluted solution on to the crystals, and agitate until dissolved. Cool, and make up to
a definite volume; the solution will have lost some ammonia from the heat and will
be acid. Ascertain the quantity of ammonia in a measured volume by distillation
with excess of normal caustic potash and titration of the distillate with standard acid.
Then titrate the excess of normal potash in the retort and calculate the citric acid
which was combined with the ammonia. Knowing the amount of citric acid and
ammonia in the measured volume, calculate the quantity of ammonia necessary to
neutralize the free citric acid. Add this quantity from a solution of ammonia of
known strength, shake, and bring to a specific gravity of 1.09. Our experience shows
that citrate solutions made neutral to litmus or corallin are generally decidedly acid.
Mr. Stillwell. The commercial chemists are constantly receiv-
ing importations of basic slag for analysis. I believe importers are
guaranteeing the available phosphoric acid by the seiving method
used at the California station, and I do not know what we are going to
do as commercial chemists unless some action is taken by the associ-
ation. I would urge that the referee recommend a method for the
determination of available phosphoric acid in basic slag as soon as
possible. Mr. McCandless spoke of the Gladding modification for
the determination of alumina ; do I understand that you get the iron
and alumina together?
Mr. McCandless. In the acetate method they are precipitated
together, but may be separated by treating with caustic soda or
potash. Most of the chemists use the Glaser method.
Mr. Stillwell. We always report the iron and alumina sepa-
rately.
Mr. McCandless. But your separation is made by means of
caustic soda or potash. We found so much alumina in potash that
the results were entirely unsatisfactory, and in the method proposed
fo'r the separation of iron and alumina it is easy to obtain chemically
pure reagents.
Mr. Hand. We have used the following method in Mississippi with
good results:
To the solution, prepared in the usual manner, add 8 grams of ammonium oxalate,
heat nearly to boiling, neutralize cautiously with diluted ammonia, and add a few
drops of dilute acetic acid to faint acid reaction. Place the beaker on a water bath
for about two hours, filter, and wash. The calcium oxalate will be practically free
of compounds of iron, alumina, and phosphoric acid. Concentrate the filtrate if nec-
essary. Heat to 70° or 80°, and destroy the ammonium oxalate by electrolysis. The
current should be passed until odor of ammonia has entirely disappeared. About 3.5
to 4 ampere hours are sufficient. The ammonium oxalate may be oxidized in an
hour and a half. So far we have used only the lighting circuit, supplying current at
220 volts.
After dissolving the mixed precipitates we have a solution free of lime, contain-
ing nothing except probably a small amount of magnesia, to interfere with the
160
precipitation of the iron and aluminum phosphates, and in ordinary rocks these may-
be thrown out at once with ammonia after the addition of about 1 gram of ammonium
phosphate. The ammonium phosphate should not be neglected. When properly
can'ied out the method has given highly concordant results. Very good results may
be secured also by destroying the oxalate by nitric acid and potassium chlorate after
evaporating almost to dryness. This procedure. howcA^er, for obvious reasons, is not
as satisfactory as the process employing the current. In presence of a considerable
amount of magnesia the iron and alumina may be precipitated as in the acetate method.
We have not had an opportunity of testing this method carefully, but propose to do
so as soon as possible.
^Ir. Veitch. The American Chemical Society has recently ap-
pointed a committee of fiye to study the determination of alumina
and iron in phosphates, and this committee ^vill be yery glad to hear
from all who are interested in this question as to their methods and
to receiye suggestions as to the lines of woxk to be followed. It
seems to me that we must consider not only the methods for the
determination of these substances, but also in what combinations
they exist. Shall we determine total iron and alumina in phos-
phates, or exclude the sulphids of these bases in the analysis ? This
is a question of some commercial importance, and we would like to
haye the benefit of the experience of the commercial and manufac-
turing chemists in considering it.
I did a great deal of work on the thiosulphate method about fiye
years ago and found it to be a most excellent method for the separa-
tion of alumina from iron in the form of phosphate. I found, howeyer,
that two precipitations would give practically theoretical results,
while one was not sufficient, usually carrying down a little iron. I
also doubt the accuracy of the method when solution is made in
hydrochloric acid and alumina innnediately precipitated without the
remoyal of silica. The effect of fluorids must also be considered, and
it is always necessar}' to wash with 5 per cent ammonium nitrate
solution, as the aluminum phosphate is h^'drolized b}" hot water,
phosphoric acid being dissolved and passing into the filtrate.
]Mr. McCaxdless. I have made some experiments in treating a
number of samples of ground phosphate rock containing lvno^ytl quan-
tities of iron sulphids yith sulphuric acid of 1.53° specific grayity, and
in no case did any of the iron sulphid go into solution. As it was not
oxidized by the sulphuric acid, it would be unfair to introduce nitric
acid into the solution, as iron sulphid does not affect the phosphate
made from such a rock. It was also found that in separating the iron
and alumina as phosphates and dividing by 2, a common practice,
very inaccurate results may be obtained.
^Ir. MooERS. I would like to say in this coimection that the Ten-
nessee phosphate contains a large amount of iron sulphid, and the
chemists at Mount Pleasant use a dilute hydrochloric acid solution —
about 1:1 strength usually. Some manufacturers haye made to me
161
the same remark made by Mr. McCandless — that the iron and alum-
ina sulphids do not affect the manufacture of phosphate.
The Assistant Secretary of Agriculture was presented by the presi-
dent of the association and addressed the convention as follows:
ADDRESS BY ASSISTANT SEOEETARY OF AaRIGULTURE.
Gentlemen: I am very glad to meet with you and to have the opportunity of
congratulating you upon the fact that your work in this association is leading to large
things; of this there is every evidence. Your peculiar work makes much possible
along the lines of the development of research, and it is also useful in the industries, in
home making, and in promoting good living. It is also educatiorial, especially in
definitely educating the people at large concerning many things that relate to farming
and the home life. You have evidence, which I know is very satisfying to you, of
your good work in some of the matters that have become the subject of National legis-
lation, especially the pure-food law, which has been largely promoted by this associa-
tion. As the plans for the execution of this law go on it is plain that the good it is
going to do will not only be economic, but also ethical. As Doctor Wiley says, it is
gt)ing to have an effect in producing business honesty in this country, which will
extend to our relations with foreign countries. The food question is one of the matters
that have been neglected in this great Republic until affairs were in rather bad con-
dition, but now we have taken hold of the matter and reorganized it. It is a great law.
I think the Department of Agriculture has reason to be proud that it has given
support to this organization and helped it in its work. Not long ago the question arose
as to whether the printing of the annual reports for other societies and this society
should continue to be the general policy of the Department, and when a statement
was made as to the relations of the Department to the work of the Association of Official
Agricultural Chemists the correctness of the policy became plain. The arrangement
you have had for the publication of your reports seems to have been very wise, and
much good has been done and many things made possible that would not have been
possible without such an arrangement.
Many difficult problems confront us in these technical questions and I know you all
are disposed to meet them in a scientific spirit. When we have any differences among
us it seems wise that we specify them, define the questions in regard to which we
differ, and apply ourselves to the solution of such definite questions by specific methods,
I am glad to be with you.
Vice-President Street reported that the following committee un
the president's address had been appointed: Messrs. Woll, Davidson,
Penny, Peters, Ross, Van Slyke, and Winton. [Mr. Winton •with-
drew from the committee subsequent upon his resignation from the
Connecticut station, and Mr. J. G. Lipman of the New Jersey station
was appointed to fill the vacancy.]
REPORT ON TANNIN.
By H. C. Reed, Referee.
Among the matters referred to the consideration of the referee for 1906 are the follow-
ing suggestions made by a committee of the American Leather Chemists' Association:
(1) That the executive committee of the Association of Official Agricultural Chem-
ists appoint a special committee on recommendations for the tannin section and that
31104— No. 105—07 11
162
the association appoint the same referee and associate referee on tannin as are appointed
by the American Leather Chemists" Association.
(2) That the Association of Official Agricultural Chemists abandon the appointment
of a referee and associate referee on tannin and consequently omit for the present any
supervision of the tannin methods.
As the reasons for these suggestions are not given in the report of last year's pro-
ceedings, the referee considers it wise to call attention to them, and in order to do this
he presents herewith a copy of the original letter embodying such reasons.
November 17, 1905.
Dr. H. W. Wiley.
Dear Sir: Confirming the conversation we had with you to-day, we herewith
present our suggestions relative, first, to the probable action of the executive com-
mittee of the Association of Official Agricultural Chemists in the matter of appointing
a special committee on recommendations for the tannin section and to the possibility
of having the Association of Official Agricultural Chemists appoint the referee and
associate referee that are appointed by our association; and, second, the inclination
of the Association of Official Agricultural Chemists to abandon the appointment of
referee and associate referee by your association and consequently omit for the present
your supervision of the tannin method.
We would request that this be done with the view of having but one method of tannin
determination, inasmuch as most of the members of the Association of Official Agri-
cultural Chemists have no interest in the work. and. in reality, the work is being done
almost entirely by members of the American Leather Chemists" Association, who can
not become voting members of the Association of Official Agricultural Chemists and
thereby protect their interests.
"We would ask that action be taken in one of the two lines proposed, and would sug-
gest that the latter proposition, being more acceptable to you, should receive the more
favorable consideration.
We feel that our interest in the tannin method is of paramount importance, repre-
senting as we do practically all of the tanning interests of this country, and that two
methods of analysis would be highly detrimental to commercial interests.
(Signed.) W. H. Teas,
H. C. Reed,
J. H. YocuM,
Committee of the American Leather Chemists' Association.
In a letter addressed to the referee dated December 16. 1905. Doctor AViley. secretary
of the Association of Official Agricultural Chemists, says:
We are very appreciative of the extent and quality of the cooperative work done by
the American Leather Chemists" Association, and shall hope that such collaboration
may be continued. Many trade interests cooperate in this way in the official work
to mutual advantage, but even a cursor^' examination of the constitution and the
objects for which our association exists makes it plain that to coalesce completely with
any trade association, however able in their investigations, would be to nullify the
authority of the methods, this authority depending largely on the fact that they are
adopted by official chemists acting from a purely disinterested standpoint. This is
without any prejudice against either the fair-mindedness or ability of the trade chem-
ists, but must necessarily be true when any special interests are involved. A neutral
or disinterested opinion carries weight. The fertilizer interests, the food men, and
the glucose manufacturers have all discussed their views, presented their methods,
etc., in the association.
The present conditions have arisen. I think, fi'om a misunderstanding of the conduct
of the association, its interpretation of the terms '•official ' ' and '"provisional " methods,
and its policy as outlined in the constitution, and I believe that we can conduct the
cooperative work on tannin in the future with mutual profit.
The referee would call attention to the fact that recommendations on tannin methods
would be presented to the executive committee of the Association of Official Agri-
cultural Chemists, approved by exactly the same body of chemists and representing
the same interests as when the recommendations are considered by the American
Leather Chemists' Association. This would mean that if trade interests entered
into the, matter at all they would have opportunity of so doing equally under the
rulings of both associations with the advantage, if an^-thing, in favor of the American
Leather Chemists' Association, since in this body the recommendations would be
163
passed upon by those acquainted with the methods, which is not altogether the case
with the executive committee of the Association of Official Agricultural Chemists.
During the past year many further reasons have appeared indicating the advisability
of the abandonment of the tannin section by the Association of. Official Agricultural
Chemists. The American Leather Chemists' Association has been rapidly increasing
in membership and has become recognized by the International Association of Leather
Trades Chemists as the representative association in tannin investigation for this
country. This is indicated in all the recent utterances of the Collegium, the official
organ of the International Association of Leather Trades Chemists, where the American
Leather Chemists' Association is now alone referred to and not the Association of
Official Agricultural Chemists.
The past year has witnessed the publication of a monthly journal by the American
Leather Chemists' Association, devoted to the interests of tannin work, and it is
respectfully pointed out that such a desirable attainment could not result under the
rulings of the Association of Official Agricultural Chemists.
Attention is also called again to the fact that, as stated in the letter of the committee
of the American Leather Chemists' Association of November 17, 1905, members of
the American Leather Chemists' Association can not become voting members of the
Association of Official Agricultural Chemists and thereby protect their interests.
Does it not seem quite unfair that recommendations made by a referee and backed
by an association actively engaged in the work should be passed upon by a committee
of which possibly only one member is by experience acquainted with the why and the
wherefore of such recommendations? Moreover, the referee understands the impos-
sibility, under the rulings of the Association of Official Agricultural Chemists, of the
appointment of a committee on recommendations from other than active members
of t"he association.
The advantage of having but one method of tannin determination is quite apparent,
and with this in mind the referee determined, after mature deliberation, that it
would be exceedingly unwise for him to submit a report upon tannin this year, which
might conflict with the report of the referee of the American Leather Chemists' Asso-
ciation. Moreover, it is exceedingly doubtful whether more than a mere handful
of analysts would have agreed to assist in the work of the Association of Official
Agricultural Chemists under the circumstances.
The appreciation of the good offices of the Association of Official Agricultural Chem-
ists is fully recognized, especially by the older members of the American Leather
Chemists' Association, who have been referees and collaborated in the tannin work
for the Association of Official Agricultural Chemists in the past, but the scope of
work now embraced has become so large as to make the handling of it awkward
under the rulings of the Association of Official Agricultural Chemists. In witness
of this the referee quotes the following from Bulletin 99, United States Department of
Agriculture, page 122.
[Note by the Editor. — The detailecl reports of the several committees have been
printed in full in the Shoe and Leather Reporter for November and December,
1905, and January, 1906, and abstracts of the report have appeared in Hide and
Leather and in Leather Manufacturer for November and December, 1905. It is
regretted that lack of space prevents the publication of the details of this valuable
and exhaustive report in these proceedings, but it is only possible to present the
outline of work and the recommendations and proposed official method as offered
by the referee for ratification and final adoption in 1906.]
It is respectfully suggested that if "lack of space"^' prevents the publication of the
referee's report on tannin then surely has the work of the tannin section outgrown
the capacity of its being cared for by the Association of Official Agricultural Chemists.
However, the referee would, in closing, urge upon your body the advisability of
adopting the recommendations embodied in the report of the referee of the American
Leather Chemists' Association for this year, provided that it is not considered advisable
to -abandon the tannin section of the Association of Official Agricultural Chemists.
164
He feels, however, very strongly, and so recommends, that the tannin section of
the Association of Official Agrictiltural Chemists abandon the appointment of referee
and associate referee and omit the supervision of the tannin method.
^Ir. TTiLEY. It is onlv fair that a statement of the case should be
made from the agricultural chemist's point of view. The production
of leather is imdoubtedlv an amcultural industrv of the first mao^ni-
tilde, and tannins, considered both in connection with forestry and
with leather making, are essentially agricultural products. TTe have
received a great deal of help from the technical chemists engaged in
the leather trade, of which we are very appreciative, and we have no
doubt of their desire to do the rio:ht thins:, but since thev have thought
best to disassociate themselves entirely from the Avork of this asso-
ciation and recommend that we discontinue the work, I am of the
opinion that it is our duty as official agricultural chemists to continue
independently the work of establishing official methods for the anal-
ysis of tamiins. using such information from the reports of the Amer-
ican Leather Chemists' Association as enable us to be perfectly fair
to the trade interests. T\liile we should prefer that there should be
complete unity of action as to methods. I. for one. would not feel
that I could support the recommendation of the referee in this case.
Mr. Yeitch. In addition to what Mr. Wiley has said. I would call
attention to the fact that the time is coming when the examination
of tanning materials will be of much greater interest to the associa-
tion than formerly, as our raw tannin materials are decreasing very
rapidly and increasing in price. I think, therefore, the groving of
tannin material will become a profitable agricultural industry and
the investigations looking to the discovery of new and available
tannins and the increase of the tamiin content in those materials
which are now known aoU constitute an important line of work.
Under the new food and drugs act there are a number of medicinal
taim.in compounds which, while not of great importance, must be
considered. I should dislike to see the association give up this line
of work, as it is more likely to grow in unportance than to decrease.
The Peesidext. I will ask Mr. Yeitch. who presented the report of
the referee, to communicate this matter to the chairman of Committee
B, who is absent.
^Ir. TTiLEY. I hope the committee will consider what has just been
said and that they will recommend that this association continue to
conduct this line of work.
The referee on insecticides. Mr. George E. Colby, was not present,
and reported by letter to the secretary that no cooperative work had
been accomplished. The follo^^ing paper was submitted by the
referee, together with the recommendation that the methods pre-
viously studied be further investigated, with a view to their adoption
as official:
165
DETEEMINATION OF KEEOSENE IN KEEOSENE EMULSION BY THE
OENTEIFUGAL METHOD.
By G. E. Colby.
The following preliminary results on kerosene emulsion are submitted in order that
others may consider the subject and an official method be developed for the examina-
tion of this class of materials.
The method followed in the work here reported is as follows:
Six, 9, or even 18 grams, according to the strength of kerosene emulsion, are weighed
and measured in cubic centimeters at 15.5° C. into a Babcock cream bottle, graduated
to 35 or 50 per cent. To this add 3 or 4 cubic centimeters of strong sulphuric acid and
twirl one minute in a Babcock machine; then add cold water and twirl again one
minute; finally add water sufficient to bring the water to or above the zero mark in
the neck of the Babcock bottle; read the cubic centimeters of kerosene obtained at
15.5° C. Calculate volume percentage of kerosene on cubic centimeters of elmusion
used in the test.
Results on kerosene emulsions (sp. gr. 0.7820 at 15.5° C.) by the centrifugal method.
Description of sample.
Actual
kerosene
used. a
Kerosene
(uncol-
ored) ob-
tained by
Babcock
method.
I. Kerosene emulsion 6 ...
66.6
28.5
• 8.6
85.7
14.2
f 66.8
i a66.5
28.7
la. Sameemulsion diluted 2.33 times (i.e., 3 gallons plus 4 gallons water)
8.7
II. Petroleum emulsion c
a 85. 7
Ila. Same emulsion diluted to 6 times original volume
14.3
a Specific gravity of kerosene at 15. 5° C. determined and found to be 0. 7820.
h Hubbard-Riley formula, p. 155, Spraying of Plants, Lodeman.
c Penny's formula No. 1, p. 23, Delaware Agr. Exper. Sta. Bui 75.
METHODS OF ANALYSIS OF LEAD AESENATE.
By J. K. Haywood.
Methods of analysis of a large number of classes of insecticides -have been gathered
together and comparatively studied during the past six or seven years by the Association
of OflScial Agricultural Chemists, but in this time no methods have been even suggested
for a very important class of insecticides, i. e., lead arsenates which are now manufac-
tured and sold by a number of firms in this country.
Lead arsenate is usually prepared by the action of lead acetate on disodium arsenate.
By studying the resulting lead arsenate it was found that the following reaction was
followed :
3 Pb(C2H30.2),+ 2Na2HAs04= Pb3(As04)2+ 4Na(C2H302)+ 2H(C2H302).
In case the crystallized varieties of lead acetate and disodium arsenate as they
usually appear on the market are used the reaction is as follows:
3Pb(C2H302)2. 3H20+2Na2H.A.s04. TH.O^ Pb3(As04),,+ 4Na C2H3O2+
2HC2H3O2+23H2O.
However, some manufacturers prepare lead arsenate l:)y the action of lead nitrate
on disodium arsenate. A study of the resulting lead arsenate shows that the following
reaction is. followed in the main :
Pb(N03)2+ Na2nAs04= PbH AsO^-f 2NaN03,
166
althougli a veiy slight variation of the resulting compound from the theoretical compo-
sition of lead hydrogen arsenate would suggest that some unknown secondar\" reaction
takes place to a small extent.
In case the cr^-stallized variety of disodium arsenate as it appears usually on the
market is used the following reaction is followed:
Pb(N03)2+Ka2HAsO . 7H,0= PbHAs04+ 2XaX03+ TH.O.
It will thus be seen that if the resulting lead arsenate were washed entirely clean of
all impmities by the manufactm'ers it would be necessary to de^^.se methods of
determining only total lead oxid, total arsenic bxid, moistm-e, and soluble arsenic oxid
(representing the solubility of the lead arsenate itself). However, the resulting lead
arsenate is not washed perfectly clean of all impurities by the manufacturer, but is
washed perhaps only once or twice or not at all and the remaining mother liquor
removed by a filter press, so that in commercial samples sodium acetate or sodiimi
nitrate are practically always present. Besides this a slight excess of lead acetate or
lead nitrate used in making the lead arsenate will also be present, on account of insuffi-
cient washing, and in certain cases where sufficient soluble'lead salt was not added to
precipitate all arsenic a small amount of sodiiun arsenate might be present. This
latter condition, however, has never been actually observed by the author in com-
mercial samples.
It will thus be seen that in commercial samples not only is it necessarj^ to determine
moistm-e, total lead oxid, total arsenic oxid, and soluble arsenic oxid, but it is also
necessary to determine soluble lead oxid and total soluble solids, which latter may,
in a broad sense, be considered as the impm-ities present in the concentrated sample.
It is the habit of the writer to state the results of the complete analysis of a sample
of lead arsenate in the following way:
Per cent.
Moisture
Total arsenic oxid
Total lead oxid
Soluble impurities i exclusive of soluble lead oxid and ar-
senic oxid)
Total
Soluble arsenic oxid '
Soluble lead oxid
Methods of Analysis.
general directions.
In case the sample is in the form of a paste, as it usually is, dry the whole of it to
constant weight at the temperature of boiling water and calculate the result as total
moisture. Grind the dry sample (which will gain a small amount of moisture by
so doing) to a fine powder and determine the various constituents as follows:
Moisture. — ^Weigh out 2 grams of the sample and heat in the water bath for eight hom's
or in the hot-air bath at 110° C. for five to six hours or till constant weight is obtained.
Total lead oxid. — Dissolve 2 gi-ams of the sample in about 80 cc of water and 15 cc
of concentrated nitric acid on the steam bath; transfer the solution to a 250 cc flask
and make up to the mark. To 50 cc of the solution add 3 cc of concentrated sulphm-ic
acid, evaporate on the steam bath to a su'upy consistency and then on the hot plate till
• white fumes appear and all nitric acid has been given off. Add 50 cc of water and
100 cc of 95 per cent alcohol. Let stand for several hours and filter off supernatant
liquid, wash about ten times with acidified alcohol (water 100 parts, 95 per cent
alcohol 200 parts, and concentrated sulphmic acid 3 parts ) and then with 95 per cent
alcohol till fi-ee of sulphuric acid. Dry. remove as much as possible of the precipitate
from the paper into a weighed crucible, and ignite at low red heat. Bm'n the paper
in a separate porcelain crucible and treat the residue ffi-st with a little nitric acid,
which is afterwards evaporated off, and then with a drop or two of sulphuric acid.
Ignite, weigh, and add this weight to the weight of the precipitate previously removed
from the paper for amount of the lead sulphate.
167
In the above method an attempt was first made to precipitate the lead alone In a
sulphuric acid solution without the addition of alcohol, but it was found that some
of the lead sulphate was dissolved and low results were obtained. It was therefore
thought best to add twice the volume of alcohol to the solution and acidify with
sulphuric acid. A test was made to determine whether the arsenic in solution would
be precipitated on the addition of this amount of alc.ohol, and the results showed
it would not. The acid mixture of alcohol was used in washing the precipitate of
lead sulphate free of arsenic for fear of precipitating this element by washing with
95 per cent alcohol alone. The 95 per cent alcohol was finally used for washing to
eliminate the sulphuric acid.
Total arsenic oxid. — Transfer 100 cc of the nitric acid solution of the sample, pre-
pared as in the above determination of lead, to a porcelain dish, add 6 cc of concen-
trated sulphuric acid, evaporate to a sirupy consistency on water bath and then on
hot plate to the appearance of white fumes of sulphuric acid. Wash into a 100 cc flask
with water, make up to mark, filter through dry filters, and use 50 cc aliquot for further
work. Transfer this to an Erlenmeyer flask of 400 cc capacity, add 4 cc of concen-
trated sulphuric acid and 1 gram of potassium iodid, dilute to about 100 cc and boil
until the volume is reduced to about 40 cc. Cool the solution under running water,
dilute to about 300 cc, and exactly use up the iodin set free and still remaining in
solution with a few drops of approximately tenth-normal sodium thiosulphate.
Wash the mixture with a large beaker, make alkaline with sodium carbonate, and
slightly acidify with dilute sulphuric acid; then make alkaline again with an excess
of sodium bicarbonate. Titrate the solution with a twentieth-normal iodin solution
to the appearance of a blue color, using starch as indicator.
The above method for arsenic oxid is a modification of the method of Gooch and
Browning a and was applied to this class of goods both on account of its simplicity
and because the same standard iodin solution used in determining arsenic in Paris
green and London purple could also be used here. The original treatment with
sulphuric acid is of course to eliminate lead. More sulphuric acid is added because
it was found by Gooch and Browning that the potassium iodid acts best on the arsenic
in sulphuric acid of a certain strength. The potassium iodid reduces the arsenic oxid
to arsenious oxid and iodin is set free. This reaction is not complete unless the solu-
tion is boiled. The solution is boiled until its volume is reduced to about 40 cc.
At this point nearly all the iodin has been volatilized, but a small amount is still
present. This is used up exactly by dilute thiosulphate solution and the mixture is
made alkaline so that the arsenic, which is now all in the form of arsenious oxid may
be titrated with standard iodin to the form of arsenic oxid.
Water-soluble lead oxid. — Place 2 grams of the lead arsenate in a flask with 2,000 cc
of carbon dioxid free water and let stand ten days, shaking eight times a day. Filter
through a dry filter and use aliquots of this for determining soluble lead and arsenic
oxids and soluble solids; determine lead as described above for total lead, using the
same relative proportions of sulphuric acid, water, and alcohol, but keeping the volume
as small as possible.
Water-soluble arsenic oxid. — For this determination use 200 to 400 cc of the water
extract obtained under the determination of soluble lead oxid. Add 0.5 cc of sul-
phuric acid and evaporate it to a sirupy consistency, then heat on hot plate to appear-
ance of white fumes. Add a very small amount of water and filter off lead through
the very smallest filter paper, using as little wash water as possible. Place this fil-
trate in an Erlenmeyer flask, and determine arsenic as described above for total
arsenic oxid, using the same amount of reagents and the same dilutions.
Soluble solids or impiirities. — Evaporate 200 cc of the w^ater extract obtained above
to dryness in a weighed platinum dish, dry to constant weight at the temperature of
the boiling water bath, and weigh. The soluble solids so obtained represent principally
any sodium acetate or sodium nitrate present, with a very small quantity, perhaps, of
lead acetate or nitrate and some soluble arsenic, probably in the form of lead arsenate.
a Amer. J. Sci., 1890, 40: 66.
168
Analytical Results.
To test these metliods two samples of lead ai'senate were prepared fi'om lead acetate
and sodium arsenate and two from lead nitrate and sodium arsenate. All the soluble
salts were washed out. the samples in fact being made as chemically pure as possible.
Since these soluble salts were removed a determination of soluble solids was useless.
Following are the results obtained by two different workers, using enthely different
standard solutions.
Results obtained on lead arsenate by the proposed methods.
Lead arsenate, chemically pure.
T)ptPTTniTifltinn<?
Prepared from lead acetate.
Prepared from lead nitrate.
Sample I.
Sample II.
Theory for
lead arse-
nate.
Sample I.
Sample II.
Theory for
lead acid
arsenate.
Total arsenic oxid:
Worker A
26.95
26.24
72.11
26.88
26.17
71.80
72.68
1 25. 60
} --
i 33. 56
1 32. 90
f 64. 13
\ 64. 04
2.31
3.06
32.79
1 _ .-
Worker B
32.92 j ^^-1^
f4?di e4-2«
Total lead oxid:
Worker A
Worker B
72 54
Water of constitution (by
difierence"* :
Worker A
3.35
1
Worker B
2.91 1 ^-^y
Total:
Worker A .
99.06
98.78
.85
98.68
98.85
.70
I 100.00
/ 100. 00
t 100. 00
.71
100.00 1 T„f. ^„
100.00 ( 100-0^
Worker B
Soluble arsenic:
Worker A
.63
Worker B
1
The table shows that the methods of analysis give very acceptable results. In the
case of lead arsenate prepared from lead nitrate and sodium arsenate, it is evident
that the compound lead acid arsenate (PbHAs04) is formed, and the results obtained
on analysis are very close to the theoretical ones. In the case of lead arsenate pre-
pared from lead acetate and sodium arsenate the results of the various determinations
on lead and arsenic agree with each other quite well, but they are not so close to the
theoretical results demanded for lead arsenate (Pb3(As04)2) as might be desired,
although they show that this compound is evidently formed to a large extent. Since
there is a tendency on the part of both analysts to obtain slightly higher results on
arsenic and slightly lower results on lead oxid, as well as slightly lower results on the
total' than is demanded by theory for lead arsenate, and since this tendency is not
shown in the case of the other compound lead acid arsenate, it would appear that
these variations are not due to inherent faults in the methods, but rather to the fact
that some secondary reaction takes place, resulting in the formation of some com-
pound other than lead arsenate. It is probable that this secondary compound is lead
acid arsenate, since the presence of a small amount of it would account for the low
result on lead, the high results on arsenic, and the low results on the total.
EEPORT or THE COMMITTEE ON FOOD STANDAEDS, 1906.
Mr. President and Members or the Association : Permit me to present, on behalf
of the committee on food standards, the following report of progress, covering the
work of the year past. The committee has held during this period three meetings,
namely, from November 20 to 21, 1905, at Boston, Mass.; from March 5 to 10 and
June 18 to 25, 1906, at Washington. By means of circular letters and drafts of tenta-
tive proposals the members of the association have been informed from time to
169
time of the subjects upon whose consideration the committee was engaged. It may
not be superfluous, however, to present at this time, for formal record in the pro-
ceedings of the association, a summary of the work performed.
The subjects announced for special consideration at the Boston meeting concerned
more particularly the standards for fruits and fruit products, flavoring extracts, edible
vegetable oils, and table and dairy salts; also the formulation of a description of suit-
able food containers, especially those made from tin plate, as necessarily involved,
under the existing conditions of food transportation and preservation, in the concep-
tion of a standard food in respect of freedom from contamination by injurious sub-
stances.
At the hearings representatives of certain interests engaged in the manufacture
of extracts urged the desirability of allowing some latitude in the use of synthetic
flavoring materials, in standard preparations sold under names suggesting natural
sources, the gist of the argument being that these synthetic products are chemically
identical with the natural flavoring principles and cheaper than the latter. A
number of valuable suggestions were made respecting the minimum strength of
standard preparations and the importance of latitude in the choice of the process
of manufacture by which the designated raw materials were to be converted into the
finished product. In connection with fruit products, the use of glucose in the manu-
facture of preserved fruit, fruit butters, and jellies was discussed, and also the nomen-
clature of products into which glucose entered as a partial substitute for sugar as
distinguished from those theoretically producible by the entire substitution of glucose
for added sugar.
Several manufacturers of fri^it products pressed the claims of evaporated apple cores
and parings for admission as raw materials for the manufacture of standard apple
products on the ground of the superior jelly-forming quality which the speakers
attributed to them.
The subject of food containers was considered at several hearings, at which the
inapplicability of the European standards to American conditions and difficulties
was urged, and also the tendency to early deterioration of foods packed in lightly
coated tin plate was discussed.
In addition to the subjects announced as the special topics for consideration at the
Boston meeting, hearings were given to officers of the association of manufacturers of
soda-fountain supplies, who presented their views respecting the standardization of
fruit juices, fruit sirups, crushed fruit, and other flavoring materials used in the
preparation of soda water. The ineffectiveness of simple sterilization for the preser-
vation of these products and the need for intensification of color by use of foreign
colors in these preparations were urged.
The processes employed in the manufacture of rum were presented by Dr. H. Saw-
yer, and the rum standard was discussed in the light of these processes and the exist-
ing analyses of rum. The representatives of other liquor interests presented their
views as to the nature of whisky and of the various so-called " blended " spirituous
liquors now sold to American consumers.
Representatives of large meat packing interests urged that the meat standards
already proclaimed be so amended as to specify saltpeter among the admissible con-
stituents; also, that the standard for salt include a water maximum, owing to the great
variability in the amounts of this constituent exhibited by salts shipped in bulk to
consumers buying in large quantities.
In executive session, the committee reviewed the correspondence received upon
the subjects assigned for action and spent some time in amending the tentative sched-
ules for flavoring extracts, edible vegetable oils, and fruits and fruit products, pro-
viding for further correspondence on certain points. Preliminary drafts of schedules
for ice cream and for spirituous liquors were also prepared.
170
The committee assembled in March of this year to complete its work on the stand-
ards above mentioned as the special topics of consideration at the Boston meeting.
The sessions were chiefly executive, no hearings being given, though much time was
giA^en to consultation with experts in the Department of Agriculture upon important
points. The representatives of certain trade interests having noted the occasional
presence of borax in refined salt, F. W. Woll, referee on salt, and W. D. Bigelow had
been requested to investigate the matter. The reports from their investigations upon
samples representing the chief regions of production indicated that the turmeric tests
for borax, ordinarily used in the detection of the substance as a food adulterant, are
not sufficiently delicate to reveal the traces occasionally present in refined salt.
Messrs, Bigelow and Chace presented data bearing upon the limits for sugar in fruit
preserves, jams, etc. ; Mr. Tolman, data respecting the range of variation exhibited by
the physical properties and chemical constituents of edible vegetable oils; Mr. Gore
contributed data upon the distribution of the pectins in various parts of the apple,
having a bearing upon the claims made for the cores and parings as soiuces of jelly;
Mr. Kebler made a number of valuable suggestions relative to the standards for essen-
tial oils and for the flavoring extracts prepared therefrom; Mr. Webster, chief of the
dairy division. Bureau of Animal Industry, reported upon the present status of the
term "factory-filled" as applied to refined salt. Mr. Coville, botanist of the Depart-
ment of Agriculture, revised the scientific botanical terms used in the standards, to
insure their conformity to the principles of botanical nomenclature now in vogue.
At this session the committee determined upon recommendations for standards
belonging to the schedules for fruit and fruit products, flavoring extracts, edible veg-
etable oils, and salt. These recommendations were approved by the honorable the
Secretary of Agriculture, and were proclaimed by him on March 8, 1906, as the National
standards established for the respective foods under the authority granted by act of
Congress. Since these standards have been published, as Circular No. 17, Office of
the Secretary, U. S. Department of Agriculture, and distributed to the members of
the association, no detailed comment upon them will be made in this report. It may
be noted, however, that in connection with the schedule for flavoring extracts, the
desirability was considered of fixing the standards upon the formulas of the United
States Pharmacopoeia for all the flavoring preparations described in that publication.
It was finally decided, that while the formulas there given afford the best basis for
the standards for certain flavoring preparations, in many cases the formulas have evi-
dently been prepared with reference to the use of the product for flavoring medicines
rather than foods. Each extract was therefore considered independently, and the
minimum of strength fixed in accordance with the best information obtainable. Fur-
thermore, inasmuch as the 1900 Revision Committee of the United States Pharmaco-
poeia had declared that the pharmacopoeial directions for making drug preparations were
not intended to ^erve as standards for preparations made for non-medicinal uses, it
was deemed best to accompany the standards recommended by the committee for
flavoring extracts, with the complementary declaration that the flavoring extracts
described are intended solely for food purposes and are not to be confounded with sim-
ilar preparations described in the Pharmacopoeia for medicinal purposes.
At the March session tentative drafts of schedules for ice cream, vegetables and xege-
table products, tea and coffee, prepared mustard, and malt liquors were formulated for
submission to the members of this association and to the trades concerned, for criticism
and suggestion; a revised draft of a description of suitable containers for holding pre-
served foods was also included. The subject of standards for spirituous liquors was
given considerable study in the light of existing analyses and manufacturing processes.
Finally, for the purpose of organizing the work upon cattle food standards, which had
been committed by the association at the 1905 meeting to this committee, it was
ordered that certain members of the association be requested to collaborate with the
171
committee, as referees on particular groups of cattle foods, and to present tentative
standards for the corresponding products.
To guide the referees, 'it was ordered that the standards to be proposed should —
(a) Be minumuin standards, similar in purpose to those already proclaimed for
human food.
(b) Conform in form of expression to those for human food.
(c) Be based upon data representing American products; that is, that they sliould
be based upon a careful study of American raw materials, manufacturing methods,
and nomenclature of manufactured products.
It was also decided that —
(d) Before the adoption of any schedule of standards, the trade be fully consulted.
(e) So far as they apply, the standards already adopted for grains and grain products
in the schedules for human food be adopted for the corresponding cattle foods.
Considerable difficulty has been experienced by the writer in enlisting the aid
desired, owing to the tremendous pressure upon the time and energies of our experi-
ment station workers. The list of referees, so far as completed, is as follows:
Committee on buckwheat, millet, sorghum, dhurra, their products: Mr. John P.
Street, of New Jersey; Mr. Frank Fuller, of Pennsylvania; Mr. L. H. Merrill, of Maine.
Committee on wheat, rye, oats and spelt, and their products: Mr. Harry Snyder, of
Minnesota; Mr. Charles H. Jones, of Vermont.
Committee on corn and its products: Mr. M. A. Scovell, Kentucky; Mr. H. C.
Midsall, Iowa; Mr. J. B. Lindsey, Massachusetts.
Committee on linseed products and other oil cakes and oil meals, except cotton-
seed meal and maize corn: Mr. H. A. Weber, of Ohio; Mr. E. F. Ladd, of North Dakota.
Committee on cotton seed and.rice products: Mr. B. B. Ross, of Alabama; Mr. B. W.
Kilgore, of North Carolina; Mr. G. S. Fraps, of Texas.
Committee on barley and brewer's and distiller's grains: Mr. F. W. Woll, of Wis-
consin.
Committee on beet pulp and other wastes from sugar manufacture, molasses grains,
also leguminous seeds and seed products used for cattle foods: Mr. J. E. Halligan, of
Louisiana.
During the interval between the March and June meetings, the tentative schedules
*for ice creams, vegetables and vegetable products, prepared mustard, cocoa butter, tea
and coffee, and malt liquors were distributed for criticism to the members of the asso-
ciation and to the several trade interests concerned; also, the tentative description of
suitable vessels for holding preserved food products was sent to a list of the principal
manufacturers of tin plate and tinware, prepared for the use of the committee through
the courtesy of the Secretary of Commerce and Labor.
At the June meeting the correspondence respecting the tentative schedules above
mentioned was carefully reviewed. Hearings were given to a number of interests.
Upon the subject of food containers representatives of the tin-plate manufacturers,
of the American Tin Can Company, and of vegetable packers discussed the question
of standard containers from their respective points of view. The committee finally
decided to extend the description so as to include the metallic foils used in wrapping
food products, to designate the quality of lin plate suitable for making containers for
certain classes of foods known to corrode defective plate, and to make certain specifi-
cations for the lacquered cans now coming rapidly into use as containers for colored
fruits and vegetables, such as strawberries and beets.
Mr. Sawyer requested a somewhat broader standard for rum than he had earlier
recommended, in view of certain experiments now in progress at the distillery where
he is engaged.
A representative of the makers and importers of the terpeneless oils of lemon and
orange urged a widening of the standards for the terpeneless extracts of lemon and
orange, to permit the use of these oils in the preparation of the corresponding extracts.
This change was later made.
Representatives of the wine industry in the eastern United States urged a revision
of the standards for wine and sugar wine, so that the use of a limited amount of sugar
172
or sugar sirup Tvitli the must might be inchided as a legitimate cellar treatment for
wine, the low saccharine content of the must of the gi-apes of this region making it
difficult to manufacture a satisfactory wine without the use of sugar, and the manu-
factm'ei-s objecting to the name "sugar wine" as suggestive of an extreme use of sugar
for such wines. In view of the conditions of the wine trade in America, especially
as they appeal to the consumer, it was decided that the broadening of the standard
for wine would not be wise: but to relieve the makers of the possible implication
attaching to the term "sugar wine," the term "modified wine" was adopted as a
designation for wines in preparing which a limited quantity of sugar or sugar sirup
has been used.
A communication was received fi'om the Italian ministry of agiicultiu'e, Chevalier
Rossati urging that the maximum limit for volatile acids in wine be raised, since
the limits in the standard established exclude an important group of Italian wines
of good quality. It was found, however, that our official method for determining
this constituent gives lower results than the method usually adopted in Italy, and
hence, that the anticipated difficulty in the entry of ^wine of good quality would
probably not occur with frequency.
Representatives of the manufacturers of vegetable products, especially canned
com and catsup, discussed the use of sugar and saccharine in the former and ben-
zoate of soda in the latter product, and Mr. Alexander Leckey, of London, England,
m^ged a more liberal standard for cocoa prepared by the so-called " Dutch process,"
and the permission to import and sell this cocoa preparation under the name "soluble
cocoa."
Finally, a committee from the United States Brewers' Association m'ged certain
modifications of the proposed standards for malt liquors. Especial objection was
made to the establishing of a separate standard for malt beer, to the specification of
a minimum storage period for lager beer, and to the exclusion of brewers' sugars from
the raw material used in the brewing of ale, porter, and stout. The manager of the
Corn Products Company urged, in addition, that "refined grits," a corn-starch product,
should be named, as well as unmalted cereals, in the list of i-aw materials for the
preparation of beer.
The tentative schedules published in May were then reviewed in the light of the
various criticisms received, were somewhat amended, and finally recommended for
proclamation by the Secretary of Agricultui'e.
Much of the time at the sessions of this meeting was spent in a revision of the stand-
ards earlier proclaimed, as the drift of legislative action in April and May made it
uncertain whether the authorization given the Secretar^^ of Agriculture to fix stand-
ards would be continued after June 30, 1906. It was therefore deemed important
that such amendments as seemed desirable in view of information gained since the
publication of the several standards be made while the power to amend unques-
tionably remained.
A number of minor changes were accordingly made in the standards earlier pub-
lished. The revised schedules were issued as Circular 19 of the Office of the Secretary,
superseding and supplemental to those published in Cu'culars Xos. 13 and 17. As the
revised standards have been sent to the members of the association, no extended state-
ment concerning the amendments is necessary in this report. Your attention is, how-
ever, briefly called to the insertion of a standard for cold-storage meat, as distinct from
fresh meat; to an extension of the limit for the colostral period in the milk standard; a
modification of the fat standards for condensed milks; a change in the cheese standards
to bring them more closely into conformity with the definitions in the act of June 6, 1896 ;
the raising of the water maximum for oatmeal; dropping the standard for so-called
whole-wheat flour, because the name is misleading: the adoption of a general defini-
tion for fruits; the specification of limits of composition for maple sugar and sirup on
the basis of investigations made and data compiled by the Bui'eau of Chemistry for the
173
use of the cominittec; the dropping of the standards for corn and ghicose sirups, earlier
recommended to conform to the legal definitions of a few States, the standards being
dropped because they did not fully conform to the principles of standardization
adopted by the committee; the modification of the standard for candy, to bring it into
conformity with the definition in the food and drugs act; the insertion of a standard
for cottonseed oil stearin, and the change of the limits of acidity for wine vinegar
and of solids for malt vinegar, in the light of more comprehensive data at the com-
mand of the committee.
It was considered desirable to give further consideration to the schedule for malt
liquors, so that action upon these standards was postponed.
It should be noted in relation to the meetings of March and June, that the committee
was favored by the collaboration of Mr. Elton Fulmer at the former meeting and Mr.
Richard Fischer at the latter, these gentlemen having been especially commissioned
as experts by the Secretary of Agriculture, so that the National Association of State
Dairy and Food Departments, whose cordial cooperation is important in securing the
unification of State standards, might be brought more closely into touch with the cur-
rent work of this committee. These collaborators joined in each instance in the recom-
mendations made to the Secretary of Agriculture.
It is not possible at this time to present to you the current work on cattle-food stand-
ards, but the work is steadily progressing, recommendations having been received from
the referees on buckwheat and maize products, animal products, sugar-beet pulp, and
molasses grains, and other recommendations are promised for the near future. It is our
hope that the committee may bfegin its part of the work of comparing and unifying these
standards at a meeting to be held during the coming month.
In closing, I desire to voice the appreciation of the committee for the cordial interest
and cooperation of the members of the association, and especially for the steady support
and valuable counsel of the honorable the Secretary of Agriculture, wi'th whom the
committee has been called to collaborate.
On behalf of the committee.
William Frear, Chairman.
On motion the report of the committee was accepted and the com-
mittee continued.
Mr. Wiley. I wish to emphasize the importance of the cooperation
now existing between the food standards committee of this association
and that of the National Association of State Dairy and Food Officials,
the latter committee including Messrs. Jenkins, Scovell, Fischer, Bar-
nard, and Fulmer. These committees now meet as one under the
authority of the act authorizing the Secretary of Agriculture to col-
laborate ^Svith the Association of Official Agricultural Chemists and
such other experts as he may deem necessary to ascertain the purity of
food products." While some degree of cooperation has existed in the
past the complete union of these two committees can not but be
extremely beneficial to the work both in the formulation of the stand-
ards and in securing the adoption of the standards agreed upon in
States where no legislation exists to the contrary, thus bringing the
State and Federal regulations into harmony.
174
EEPOET or THE COMMITTEE OlS EEETILIZEE LEGISLATIO]^.
By H. W. Wiley. Chairman.
During the past year the chairman has had little opportunity either to study the
subject of a Xational fertilizer law or to communicate on the subject with his fellow
members. He has, however, acting under the instructions of the committee, had
two conferences with Mr. W. H. Bowker, representing the manufacturing interests.
It was the opinion of the committee that any legislation which might be framed should,
if possible, receive the support not only of the State chemists charged with the inspec-
tion of fertilizer materials, but also of the agricultm-al manufacturing interests. It is
believed by yom- committee that it is possible to fi-ame a measm-e so ethical in its
principles and so general in its application as to merit and secm'e the support of all
interested parties. To this end the chairman submitted to Mr. Bowker a rough draft
of a proposed measure drawn for the piu'pose of regulating interetate commerce in
fertilizing materials, and asked him also to submit a tentative measure which the
committee might consider. In answer to this request a reply was received from Mr.
Bowker under date of October 18. 1906, which read in part as follows:
You will remember that you asked me in the spring to prepare a tentative bill
acceptable, in my opinion, to the fertilizer interests, and I have been working on such
a measure, but it does not altogether suit me, and. what is more to the point. I could
not offer it as a measure which, in my opinion, would be acceptable at this time, to the
fertilizer manufacturers. I am afraid that there are as many minds as there are manu-
facturers, and that we can never get together until we are face to face with proposed
legislation. \\Tien I last saw you in Xew York you will remember that you expressed
the opinion that the matter should go over for a year or two, and I concurred. In
that event you and your associates can study the matter and come to some agree-
ment, and meantime I will see what can be done with the manufacturers.
Ami:hing which I might present now would be only my personal measure , and * * *
I am not willing to stand sponsor for any measm'e without fm'ther consultation. I
think, therefore, the whole subject should be left as it was last fall, and you will recall
that I suggested in my remarks that the official chemists should first get together and
agi'ee upon some measure and then submit it to the manufacturers, who. when they
are face to face with such a proposition, will, no doubt, appoint a committee to meet
a committee of yom* association. * * * i think it would be premature for me to
present a measm-e now and perhaps create discord, whereas, in the interests of the
whole, we want harmony and cooperation. * * *
After a conference with Mr. Bowker the chairman addressed the following letter to
hun on October 29, 1906:
I am very content to let the matter rest awhile until there is more of a crystallization
of sentiment regarding it. I think in the end there should be no difficulty in an
agi'eement between oiu* association and the association of manufactm'ers. I think this
for two reasons; first, the association of manufacturers is made up of honest men, who
want to give the, equivalent of the prices for the goods which they fm-nish. They are
men who are perfectly willing to mark their goods so as to show Iheh' character. On
the other hand, our association is composed of honorable men who wish to secure the
protection above mentioned for their clients and for the farmers. There are many
details respecting the character of labeling and the character of the guarantee, the
discussion of which should be approached in a friendly spirit. * * *
During the past year, as you all know, important measures regulating interstate
commerce in foods and drugs were pending before the Congress of the United States.
To secure successful legislation on these measures required the united efforts of all
interested parties. It appeared to your committee that it would be unwise at such a
time to distract the attention of Congress and of the friends of legislation by asking
for the consideration of another measure of a similar character. If the principle of
the control of interstate commerce in products of this kind is successfully maintained
in the administration of the food law and is supported by the courts, it will be less
difficult in the future to secure the needed legislation of a National character affecting
interstate commerce in fertilizers.
175
It is considered to be the duty of this committee to call attention also to the reasons
which occur to many for considering the proposed legislation inopportune or unneces-
sary. One of the chief of these reasons is the fear that National legislation regulating
interstate commerce in fertilizers may interfere with existing State legislation, espe-
cially in regard to fees for inspection or licenses. These views were well summarized
in the National Stockman and Farmer for November 8, 1906, from which the following
excerpt is taken:
We do not favor any such legislation at this time for several reasons, a few of which
will be mentioned. First, there is no necessity for a National fertilizer law and no
demand for it l)y the people in whose interest such legislation should be enacted.
There is no necessity for it, because twenty-seven States have fertilizer laws of their
own — every State in which fertilizers are used in any quantity. These laws are
enforced by State officials, with the aid of the experiment stations in some cases.
As a rule they are adapted to the needs of the States and are as well enforced as other
State laws — better perhaps than a National law could be enforced. That they are
satisfactory to the interests they are intended to protect is evident from the lack of
complaint by those interests, who are responsible for their existence. If there is any
demand from farmers for a National fertilizer law, we have never heard of it. No
agricultural or live-stock association, no agricultural journal, that we know of, has
voiced such a demand. The chemists and the manufacturers of fertilizers seem to
be the sole power behind the movement, and they are not the people in whose interest
fertilizer legislation should be brought forward.
It is therefore recommended that the committee be continued, but with instruc-
tions to secure, if possible, the collaboration of the great fertilizing interests in the
perfection of a measure which in the future may be submitted to the Congress of the
United States with the joint approval of the manufacturing interests and the Associa-
tion of Official Agricultural Chemists.
H. W. Wiley, Chairman.
H. B. McDonnell.
B. B. Ross.
On motion by Mr. Frear the report was ordered to be received by
the association, and the committee continued.
Mr. McCandless. The Georgia fertihzer law is the one agreed
upon at the Hot Springs convention in 1900 by the Cotton Associa-
tion of Commissioners of Agriculture and the fertilizer manufac-
turers. A bill was drafted by the commissioners, with the advice
of the State chemists, and was discussed for several days, section by
section, with the result that a measure acceptable both to the manu-
facturers and the commissioners was finally adopted." This measure
was immediately enacted into a law in Georgia. Alabama and
Tennessee followed our example, and, I believe, Mississippi. We
have been operating under this law ever since, and I believe we have
in it a nucleus for a National fertilizer law which would be acceptable
not only to the State officials, but also to the great manufacturing
interests.
Mr. Frear. It occurs to me. that there is one class of interests
which has not yet been brought into touch with this movement
whose judgment and collaboration it is desirable to have. I refer to
the officials charged mth the execution of the State fertilizer laws.
Some of us are simply chemists whose duties end with the examina-
a U. S. Dept. Agr., Burea^u of Chemistry, Bui. 81, p. 207.
176
tion of the samples and an expression of opinion as to their quahty.
All the other duties of the fertilizer control are conducted by officials
of superior rank. A certain uniformity in the requirements as to
labeling, declaration of materials, and the return that should be
made for the money expended is highly important, and the method
of securing such uniformity ought to be considered b}" all concerned.
I should like to see the scope of the committee enlarged, so as to bring
into consultation these State officials.
^Ii\ Wiley. As the fertilizer control is almost exclusively m the
hands of the agricultural experiment stations, which this association
primarily represents, I thmk the committee would naturally consult
the other officers connected with the fertilizer control without further
instructions.
^Ir. BowKER. As a manufacturer I wish tD say that while we are
not seeking this law, as the report of the committee might seem to
imply, but are satisfied to work under the present State laws, still if
it is decided that a National law will be better we will accept that
decision, provided the law is one under which we can work. I do
think that some uniform statement of analysis is very desirable;
whether a National law shall go any further than that is for the
various interests to determine. We have no wish to avoid paying
license fees under a National law, but in some States there are no
fees, and it does seem as though there should be some uniformity in
the matter. For example, it seems to me unjust that the State of
Rhode Island should charge $18 for a complete analysis when we
only pay $15 in some of the States that use many tim.es as much
fertilizer.
It seems also that there should be some change in the statement
required in regard to certain fertilizer ingredients, particularly potash.
If it is proper to give the water-soluble and available phosphoric acid,
why not give the water-soluble and available potash? The reverted
phosphoric acid is determined by a purely arbitrary method, and a
similar arbitrary method could be adopted for the determination of
available potash. In all of the organic substances used in fertilizers,
and there are many of them, there is from 0.25 to 3 per cent of potash,
for which the manufacturer gets little if any credit under the present
method. In closing, let me say again that the manufacturers do not
oppose the State laws and they will not, in my opinion, oppose a
National law, provided it is fair to all interests.
^Ir. McCandless. We have found in Georgia during the past season
considerable adulteration of the nitrogenous materials mixed with
commercial fertilizers, and I should like to present a short paper on
the subject, if it is the pleasure of the association.
177
ADULTEEATION OF OOMMEECIAL FEETILIZERS.
By J. M. McCandless, State Chemist, Georgia.
In the past few years the price of nitrogen has been rising by leaps and bounds,
and of course the temptation to use any substitute has grown continually. In the'
summer of 1905 we became suspicious that all was not right with Georgia fertilizers,
and our department sent out a circular letter to the fertilizer trade, warning them
against such materials as dried muck or peat and materials containing cyanogen
compounds, samples of which had fortunately been obtained. At the beginning of
the season of 1905-6 a systematic examination was begun of every sample to determine
the nature of the ammoniating material. It was not long before leather and also
Prussian blue or ferrocyanid of iron were discovered in the samples of mixed fer-
tilizers. Our law prohibits the use of leather in fertilizers, unless it is registered at
the department of agriculture as a source of ammonia and satisfactory evidence
submitted to show that it has been so treated as to render it available. In the cases
referred to no such registration or evidence had been submitted, but the samples on
analysis by the pepsin-hydrochloric acid method, and by the alkaline permanganate
method, showed an availability of about 75 per cent, so that the violation of the law
was technical rather than real, especially as the goods were sold to manufacturers
under the name of tankage, and the only action taken was to seize a car belonging to
a maker of the treated leather and ship it out of the State, at the same time publishing
the fact.
In the case of the samples containing Prussian blue or ferrocyanid of iron the
available nitrogen was low, as was to be expected, and the names of the manufacturers
have been published in our annual bulletin. This Prussian blue has been put on
the market under the names of beet-root manure, potash manure, and fillerine. An
examination of these raw materials showed them to be composed chiefly of oxid of
iron, ferrocyanid of iron, free sulphur, ammonium-sulphocyanid, ammonium sulphate,
and naphthalin, evidently showing them to be of gas-house origin. Although sul-
phocyanids can readily be detected in the original material, after mixing with acid
phosphate, the reaction seems to be prevented by the presence of phosphoric acid, so
that sulphocyanids can not be detected in the mixed goods by the iron test. The
Prussian blue, however, can readily be detected.
Other materials of a low nitrogen availability that are found on the market are dried
muck or peat, mora meal, and grape or tartar pomace. The systematic plan of pro-
cedure for the detection of adulterants which was adopted last season is as follows:
Prepare a dilute solution of sulphuric acid (25 cc of concentrated acid to 1 liter of
water), also a dilute solution of caustic soda made by adding 100 cc of sodium hydroxid
solution of 1.40 sp. gr. to 1 liter of water. In the examination for organic adulterants,
as leather, muck, mora meal, place 5 to 10 grams in a porcelain dish, stir with water,
and then agitate in the same way that a. gold miner treats the pulverized ore in his
pan, bringing the lighter organic matters to the surface and transferring them to a
small beaker, in which they are washed two or three times by decantation to remove
as much soluble phosphoric acid as possible. Then pour 10 or 15 cc of the dilute
sulphuric acid onto the residue and bring the whole to a boil, remove the lamp and
smell the hot solution at once. In the presence of leather its unmistakable odor will
be noted in almost every case; filter the solution and test for tannic acid by Dabney's
test.
Usually the simple neutralization of the sulphuric solution with ammonia is suffi-
cient to give the purple-wine color characteristic of leather. Further confirmation of
the presence of leather is obtained by transferring the residue on the filter to a small
beaker and boiling with the dilute solution of caustic soda; if leather be present it
31104— No. 105—07 12
178
dissolves completely, producing a deep red color. Some experiments were made
which show that an approximate determination of the percentage of leather present
may be made by comparing the color of the diluted solution with that of a standard
solution of leather dissolved at the same time and under the same conditions. If
leather be absent and muck present, no reaction takes place until treatment with the
dilute boiling caustic soda solution, when in the presence of muck a deep brown-black
solution will be obtained. Mora meal gives a strong reaction for tannic acid and might
be easily mistaken for leather, as it also gives a deep red color on treatment with
caustic soda solution. But it differs from leather in yielding a different odor on
boiling with dilute sulphuric acid and in not dissolving in the caustic soda solution
completely. On boiling a small portion of leather with soda twice, decanting, and
washing no further color will be obtained from the leather, but if mora meal be present,
the hot soda extracts a red color for three or more treatments. Under the microscope
also mora meal differs from leather.
The heavy residue from the panning to remove organic matter will contain the
Prussian blue and it may be detected by boiling with dilute caustic soda, filtering,
neutralizing with hydrochloric acid, and adding a drop or two of ferric chlorid, when a
blue color or a precipitate of Prussian blue will be obtained. If considerable organic
matter be present and the Prussian blue has been used only in small quantity, a green
solution will develop on standing a few minutes, due to the fact that the yellow color
derived from the organic matter gives a green with the blue derived from the
ferrocyanid.
It is not inappropriate to call attention to the fact that we shall afford a much more
perfect protection to the consumers of commercial fertilizers by hunting for specific
low-grade materials in mixed fertilizers than by attempting to determine the avail-
ability of the nitrogen in any given sample. For instance, suppose a mixed fertilizer
to contain in one ton 600 pounds of an ammoniating material showing an aA^ailability
of 95 per cent and 200 pounds of a worthless stuff showing an availability of only 30
per cent. The mixed goods would show an aA'ailability of 78.7 per cent and be rated
very high.
EEPOET OP THE COMMITTEE ON DEFINITIOIT OF PLAINT POOD.
Attention is called to the report of last year, published on page 197 of Bulletin 99 of
the Bureau of Chemistry. Your committee has considered the basic principles of the
definitions proposed by the committee of which Professor Barnes was the chairman
and our own, and are of the opinion that the language of a definition which would be
agreeable to all parties has not yet been invented ._ Your committee is further of the
opinion that the difference between the two apparently opposing schools is more a
matter of verbiage than of reality. The committee of the botanists defines what your
committee reports as "plant food" as "food materials." It appears that this is rather
a distinction than a difference. Nevertheless, it has appeared advisable to secure
additional information on this subject, and to this end your committee has consulted
some of the leading authorities and has made liberal extracts therefrom. The following
authorities have been examined:
(1) A Manual of Botany, by J. Reynolds Green, 1902, Vol. II, Classification and
Physiology. Green unequivocally and tersely supports the Barnes theory respecting
plant foods. This is especially brought out on page 405 of the book.
(2) Food and the Principles of Dietetics, by Robert Hutchison, Assistant to the
London Hospital, third edition, 1901. Hutchison defines food as follows: "A food
may be defined as anything which, when taken into the body, is capable either of
repairing its waste or furnishing it with material to produce heat for nervous or
muscular work."
179
(3) Standard Dictionary: "Food is that which is eaten. or drunk for nourishment;
aliment; nutriment, in a scientific sense; any substance that, being taken into
the body of animal or plant, serves, through organic action, to build up normal
structure or supply the waste of tissue; nutriment; aliment, as distinguished from
condiment.
"Plant food — anything adapted to sustain the growth of plants; the portion of
natural materials or of fertilizers that plants can assimilate."
(4) Webster's Dictionary: "Food is what is fed upon; that which goes to support
life by being received within, and assimilated by, the organism of an animal or a plant;
nutriment; aliment; especially, what is eaten by animals for nourishment."
(5) Century Dictionary: "Food is what is eaten for nourishment; whatever supplies
nourishment to organic bodies; nutriment; aliment; victuals; provisions. 2, Any-
thing that sustains, nourishes, and augments. 3. Anything serving as material for
consumption or use."
(6) A Digest of Metabolism Experiments, by W. 0. Atwater and C. F. Langworthy,
Bulletin 45, Office of Experiment Stations, 1898. These authorities also define food
as that which is taken into the body.
(7) Farmers' Bulletin 142, U. S. Department of Agriculture, Principles of Nutrition,
by W. O. Atwater, 1902. This also supports the theory that foods are only the mate-
rials entering the body. He says, "Food is that which taken into the body builds
tissues or yields energy."
(8) Manual of Cattle Feeding, by Henry P. Armsby, fifth edition, 1890. Armsby
also sustains the theory that the term foods is applied to substances entering the plant
or animal body.
(9) The Principles of Animal Nutrition, by Henry P. Armsby, 1903. This work also
sustains the theory above mentioned.
(10) Feeds and Feeding, by W. A. Henry, sixth edition, 1904, page 3. "How plants
gather food." Henry shows that the inorganic materials entering the plant are the
real food thereof.
(11) Lectures on the Physiology of Plants, by Julius von Sachs, edition of 1887,
page 282. Sachs strongly supports the theory that the food of plants is what enters
them from without.
(12) Agricultural Botany, Theoretical and Practical, by John Percival, second edi-
tion, 1902, page 201. Percival rather takes both sides of the case, saying, "Green
plants likewise need food of a similar complex nature for development and growth;
they are, however, not generally adapted to obtain compounds of this character from
their surroundings, but are able to manufacture them from inorganic compounds, such
as carbon dioxide, etc."
(13) Works of Liebig. Liebig was perhaps the original founder of the inorganic
theory of plant foods, and supports it strongly throughout all his works.
(14) The Relations of Plants to the Soil, by J. A. CI. Roux, Paris, 1900, page 3.
Definition of Plant Food. — "Every body which penetrates into the plant and which
alone or combined with other bodies helps in its nutrition constitutes a food. The
foods are gaseous (oxygen, carbonic acid, nitrogen); liquid (waters); and solid (organic
and mineral matters)."
(15) A Text-Book of Plant Physiology, by George James Peirce, 1903, page 43.
Peirce supports the theory that food is what is taken into the body.
(16) Lectures on the Physiology of Plants, by Sydney Howard Vines, edition of
1886, Lecture VIII, page 122. Vines strongly supports the theory which considers
foods as the substances which enter the organism.
(17) Die Landwirthschaftlichen Versuchs-Stationen, vol. 29, 1883, page 253. Arti-
cle by von Raumer recognizes mineral matters as foods of plants. Same work, 1878,
page 100, vol. 21 — a distinct recognition by Stutzer of the mineral food of plants.
180
(18) Tlie Physiology of Plants, by Wilhelm Pfeffer, vol. 1, second edition, Ewart's
translation. Recognizes the external origin of foods for plants.
(19) Lectures of Jost, 1904. In the seventh lecture, page 95, he sets forth the prin-
ciple of the inorganic character of the food of plants.
(20) How Crops Grow, Johnson, edition of 1900, pages 366-369. A strong support
of the mineral food theory of plants.
(21) The Xutrition of the Plant, by Lotiis Grandeau. Paris. 1S79, page 3. A strong
support of the mineral theory of plant food.
(22) Annales de la Science Agronomique, second series, third year, 1897, vol. 1,
page 175. Gives strong support to the mineral theorj^ of plant food.
(23) The Physiological Role of Mineral Nutrients in Plants, by Dr. Oscar Loew,
Bulletin 45, Bureau of Plant Industry. Preface by Dr. Albert F. Woods, supporting
the theory of the mineral nutrition of plants.
(24) Agriculture in Some of Its Relations with Chemistry, by F. H. Storer. seventh
edition, vol. 1, 1897. He says, "On considering the relations in which plants stand
to the air and the soil which stuTound them, the questions natiually arise. What are
the sotu^ces from which plants derive food? and. How is it that plants take in theh
food?"
(25) How Crops Feed, by Samuel TV. Johnson, edition of 1900. Complete recog-
nition of the theory of mineral plant food.
(26) Annales de la Science Agronomique, second series, 1902-3, vol. 1. Complete
exposition of the mineral theory of plant nutrition.
(27) A Text-Book of Botany, by Edward Strasbtu'ger, Fritz Xoll. H. Schenck. and
A. F. W. Schimper, 1903, page 171. "The Essential Constituents of Plant Food.''
Complete exposition of the mineral theory of plant food.
Many more authorities might have been seciu*ed, but, as will be seen, fully half of those
consulted are botanists, and all but two of the authorities consulted give tinequivocal
support to the theory that plant foods are wholly of external origin and principally
of a mineral character. It does not seem wise to prolong a discussion which turns
simply upon the meaning of a phi-ase. It is perfectly evident that our botanical
brethren who differ fi'om us in om' definitions do so only pro forma. We can not
see what difference there is betw:een the expressions "foods" and "food materials."
They mean one and the same thing. Your committee therefore recommends that we
adhere to Oiu- definition of plant foods as those substances entering the plant from
without and which are utilized in the metabolic activities, and that we adopt the
definition of our botanical brethren as food materials meaning the same thing to them
as foods do to us.
Respectfully submitted.
H. AV. Wiley, Chairman.
L. L. A'an Slyke.
E. W. ^Iagruder.
The report was accepted by the association and the committee
discontinued
The meeting adjotirned.
181
FRIDAY— AFTERNOON SESSION.
EEPOET or THE COMMITTEE ON THE TESTING OP CHEMICAL EEAGENTS.
By L. F. Kebler, Chairman.
Since the last annual meeting of the Association little work of note has been done
in testing chemical reagents. The committee of the American Chemical Society in
charge of this subject met last March and discussed ways and means for the purpose
of placing the nomenclature on a satisfactory basis and the tentative names agreed
upon were to be submitted to various manufacturers, not only in this country, but
also to foreign manufacturers, for criticism and suggestions. The impurities to be
considered in connection with a number of chemicals were also discussed.
The secretary of the committee has informed the writer that as yet very few definite
replies have been received from the manufacturers consulted. Particularly was this
true of the foreign manufacturers. In view of the fact that during the two previous
years little had been accomplished by laboratory work, it was decided to pursue
another course. The work was confined to testing the various chemicals commonly
employed in analytical work for the presence of arsenical compounds and chlorids.
The OccuRRENfiE op Arsenic in Chemical Reagents. «
The first point to be considered in undertaking this work is to ascertain what method
or methods are most suitable. Among the methods tried may be mentioned, first,
the Gutzeit test, which is based upon the reaction of arsenureted hydrogen and
mercuric chlorid. The test is usually made by placing the acidified solution of the
sample in a test tube containing metallic zinc and allowing the generated gas to pass
through a plug of cotton or asbestos containing purifying and drying agents, and then
to diffuse through a mercuric chlorid spotted filter paper placed over the mouth of the
test tube. Many modifications of this method have been proposed, but no general
agreement as to its value has apparently been reached. From our own experiments
the conclusion is drawn that this method is unreliable for quantities of arsenic less
than about ten parts per million .
The second method to be mentioned is that known as the Reinsch test and depends
for its applicability upon the deposition of arsenic from a warm hydrochloric-acid
solution upon the surface of pure copper foil in contact with the liquid. The arsenic
so deposited can then be removed from the copper foil by heating in a hard glass
tube. The arsenic is oxidized to arsenious oxid and condensed as a white sublimate
in the colder part of the tube. This method has not given satisfactory results in this
investigation.
The Marsh-Berzelius method was carefully tested and found to give the most satis-
factory and reliable results. The only objection to the method is the amount of time
necessary to operate it. Many modifications have been suggested, but the principle
of all is the same and the degree of accuracy which can be obtained is probably due
more to the care exercised by the individual worker than to the new and improved
forms of apparatus which may be used.
In the work described in this paper the ordinary Marsh-Berzelius apparatus and
the electrolytical apparatus of the form described by Thorpe b were used indiscrim-
inately throughout. No novel features were introduced in the case of either apparatus
and it therefore does not appear necessary to describe in detail the general procedures.
A number of experiments indicated that the arsenic mirror obtained by the electro-
o The experimental part of the work was done by A. Seidell, Bureau of Chemistry.
b J. Chem. Soc, 1903, 83: 974.
182
htical and Marsh-Berzelius methods are comparable in all respects and therefore
that equally satisfactor\- determinations can be made by using the one or the other
apparatus.
The chemical reagents tested in this Trork are purchased on annual contracts awarded
to different firms. It therefore happens that the chemicals reported represent only
a few manufacturers. From our general experience, however, they are fairly repre-
sentative of the quality of chemicals usually supplied to chemists. It will be seen
that while many of the samples contained practically no arsenic, there are a few
others in which its occurrence appears general. Only five of the twenty-three samples
of ammonia salts, incltiding the acetate, carbonate, chlorid. nitrate, etc.. were foimd
to contain appreciable quantities of this impurity. Two samples of ammonia alum
contained, respectively, 37.5 and 40 parts of arsenic per million. From this it appears
that with the possible exception of ammonia alum the occurrence of arsenic in ammonia
salts is imcommon, and when found its presence is probably due to accident rather
than to any inherent tendency of the product to retain arsenical compotmds.
Out of seven samples of barium salts, two showed 1 part per million of arsenic each.
Six samples of boric acid were examined, and one marked "IJ. S. P." was found to
contain 10 parts per million of arsenic. The othei-s were free from this impm'ity.
The calcium salts examined consisted of thi-ee samples of carbonate, six of the chlorid,
twelve of the ox id, and two of sidphate. Of these, arsenic was found in one of the car-
bonate samples, three of the chlorid, and in both of the sulphate samples. From this
it appears that calcium salts are more or less contaminated with arsenic and therefore
should be carefully tested when the subject of arsenic investigations are under con-
sideration.
Three otit of six samples of copper sulphate contained fairly large amotmts of arsenic.
It has been known for a number of years that glycerin is particularly apt to contain
larger or smaller quantities of arsenic. Dming recent years, however, the amounts
present in this chemical have been materially reduced; but that glycerin is not free
from this impurity is attested by the analyses made public from time to time. In
confirmation of these reports we find that of fifteen samples received from nine different
sources all except one showed distinctly measureable amounts of arsenic. The amount
found varied from a trace to 3 parts per million.
Only one of six samples of sulphuric acid showed a faint trace of arsenic. On the
other hand only one sample out of six of hj'drochloric acid was free from this impurity
although it is fi-equently sold as being arsenic free. All of the samples of nitric acid
examined indicated the presence of traces of arsenic.
The potassium salts are comparatively fi-ee from arsenic compounds. Out of twenty-
four samples examined only the permanganate and sulphite indicated the presence
of considerable quantities of arsenic. The permanganate samples were secured from
three different manufactm-ers and contained from 3 to 7 parts per million. One sample
of potassium sulpliite was found to contain 70 parts per million of arsenic.
Thii-ty-six samples of sodium salts were examined. Of these all were fairly free
from this impurity, except the sodium sulphite and sodium bromid. The amoimts
present in the sulphite varied fi-om 2 to 7 parts per million.
MixuTE Qlaxtittes or Chlorids in Chemical Reagents. «
Determined by comparison oi the silver chlorid opalescences.)
During the coui'se of the examination of large numbers of chemicals in the drug
laboratory of the Bm-eau of Chemistry it has been noticed that the ordinary qualitative
test for chlorids indicates the presence of this imptuity in a majority of the samples
examined. It therefore appeared desirable to make a more careful study of the
matter and if possible to discover the source of this freqaent occurrence of chlorids in
a By L. F. Kebler and A. Seidell.
183
chemicals, and also to form an opinion regarding the establishment of an allowable
limit for chlorids in various classes of chemicals.
The method employed for making the tests is briefly as follow^: Aqueous solutions
of weighed amounts of the chemicals to be tested are placed in 50 cc Nessler tubes,
acidified with nitric acid, and aqueous silver nitrate solution added. The silver
chlorid opalescences which result are compared with those developed in a similar
manner in a series of Nessler tubes containing known amounts of a standard thousandth-
normal hydrochloric acid solution. The comparisons were made in a room lighted
by artificial light, observing the tubes from the side by means of reflected light.
The use of a nephelometer as described by Richards and Wells ^ would, undoubtedly,
have permitted very much better readings, but unfortunately such an instrument was
not at hand. The results obtained, however, are reasonably satisfactory for the pur-
pose in view and are accurate to within 3 to 10 parts per million, depending upon the
amount of sample used for making the test. In addition the method employed is one
which can be used in any laboratory supplied with Nessler or large glass tubes. '
Before beginning the actual examination of the samples of chemicals a number of
experiments were made to ascertain to what extent the salts necessarily in the solution
when the test is applied affect the intensity of the silver chlorid opalescence. It is
stated by Wells & that concentrations of other salts in amounts less than 0.01 normal do
not interfere with the production of the opalescence. This, however, applies to read-
ings made with the very sensitive nephelometer and is no doubt a much smaller amount
of salt than would seriously affect the accuracy of the comparison made according to the
method used in the present work.
Since all tests are made in solutions acidified with nitric acid, the effect of this acid
upon the opalescence was first determined. This was done by adding gradually increas-
ing amounts of the purest nitric acid, which showed no chlorids when treated with sil-
ver nitrate, to 50 cc Nessler tubes and diluting to the mark with water; 1 cc portions of
thousandth-normal hydrochloric acid were added to each tube and then 1 cc of 10 per
cent silver nitrate solution. The opalescences produced were so nearly the same in all
the tubes that no differences could be detected by the eye. The addition of 0.5 cc of
thousandth-normal hydrochloric acid to some of the tubes produced differences which
were easily distinguishable. It therefore appears that 60 parts by volume of concen-
trated nitric acid in 100 parts by volume does not appreciably affect the silver chlorid
opalescences.
The presence of various amounts of sodium nitrate was next studied by adding
gradually increasing amounts of concentrated sodium hydroxid solution (the sodium
hydroxid being prepared from sodium and showing no chlorid when tested with silver
nitrate) to the Nessler tubes and acidifying each solution with the pure nitric acid.
One cubic centimeter of the thousandth-normal hydrochloric acid and one of the aque-
ous silver nitrate solution were added as before and the resulting opalescences proved to
be identical. In addition to this test, a series of Nessler tubes was prepared with a con-
centrated sodium nitrate solution made as above and increasing amounts of thousandth-
normal hydrochloric acid solution. Another series was prepared with water and
increasing amounts of thousandth-normal hydrochloric acid. When the silver nitrate
solution was added to the tubes of these two series it was found that the opalescences
produced in corresponding tubes were the same as nearly as the eye could judge. From
this it appears that to make the standards for judging the unknown samples it is equally
as satisfactory to use water alone as to use concentrated chlorid-free solutions of the
same salt that is under examination . " Therefore, the standard tubes which were used in
judging the amounts of chlorid in the samples enumerated in the following tables were
made by adding known amounts of the standard thousandth-normal hydrochloric acid
solution (1 cc, 2 cc, 4 cc, etc.) to Nessler tubes containing distilled water acidified with
aAmer. Chem. J., 1904, 31: 235. ^Amer. Chem. J., 1906, 35: 99.
184
a few cubic centimeters of pure nitric acid and simultaneously treating these tubes, and
the tubes containing the weighed amounts of the sample in solution, with the silver
nitrate solution un^er similar conditions of light, temperature, etc.
The weight of sample ordinarily used was 10 to 20 grams, which was thoroughly
shaken with 100 cc of water and 25 cc of the clear solution placed in the Nessler
tubes and acidified with chlorid-free nitric acid. The opalescence was produced in
all cases by the addition of 1 cc of 10 per cent silver nitrate solution.
It is to be regretted that the samples available in the laboratory at the time the exam-
ination was made were not representative of a larger number of manufacturers. There
is reason to believe, however, that they exhibit the average quality of chemical reagents
usually found in the market at the present day.
Among the first lot of chemicals examined for chlorids were the various salts of
potassium, and the results which were obtained are given in Table 1.
The abbreviations used in the tables are as follows:
Mallinckrodt Mallinckrodt Chemical Works, St. Louis, Mo.
B. and C Bullock & Crenshaw, Philadelphia, Pa.
B. and A Baker & Adamson Chemical Company, Easton, Pa.
Merck Merck & Co., Darmstadt. •
E. and A Eimer & Amend, New York.
A. H. Thomas A. H. Thomas Company, Philadelphia, Pa.
Schering Germany.
Geo. D. Feidt Philadelphia, Pa.
Kahlbaum Germany.
B. and L Bausch & Lomb Company, Rochester, N. Y.
Baker J. T. Baker & Co.
Mackall Bros Washington, D. C.
Henry Heil St. Louis, Mo.
Table 1. — Chlorids in potassium salts.
Salt.
Description on
label.
Source.
Parts per million.
Serial
No.
Potas-
sium
chlorid
Chlorin.
Acetate
C. P. crystals
do
Mallinckrodt
None.
None.
None.
30
None.
88
59
None.
None.
■59
None.
None.
None.
147
None.
None.
None.
30
726
30
380
None.
None.
None.
120
240
60
None.
60
60
240
30
90
90
180
330
224
None.
491
do
do
None.
833
do
. .do
B. and C
None.
1341
do
do
B. and A
14
do
214
do
Merck
42
712
- do
do
C. P. crystals
.. ..do
Imported by E. and A
28
821
B. and A
924
do
do
do
None.
928-
do
Merck
28
1433
do
C. P. crystals
do
B. and A
do
do
None.
do
None.
do
do
do
70
249
do
Crystals, pure
C. P. crystals
Merck
None.
511
do
Mallinckrodt
None.
790
do
Diamond Soda Works
None.
860
Citrate
Chemically pure . . .
B. and A
14
968
do..
E. and A
345
1173
do
Chemically pure...
Yellow
A. H. Thomas
14
215
Merck
171
325
do
.do
Chemically pure . . .
do
Mallinckrodt
None.
425
do
None.
573
do
do
do
None.
1294
do.. .
do
B. and A
56
1427
do
do
do
112
do
do
...do
28
211
Bichromate
do
Crystals, H. P
Chemically pure...
do
Merck
None.
1282
B. and A . .
28
1295
do
do
28
296
. do.
.do
do
112
1441
do
do
do
14
1484
do
do
do
42
1000
do
do
.do
42
do
do
do
84
596
Hydro xid
Pure lumps
Pure by alcohol
Schering
157
813
do
B. and A
106
185
Table 1. — Chlorids in potassium salts — Continued
Salt.
Description on
label.
Source.
Parts per million.
Serial
No.
Potas-
sium
chlorid.
Chlorin.
225
Nitrate
Merck
None.
Trace.
30
30
None.
None.
1,130
4,520
None.
59
42
None.
Trace.
Trace.
6
None.
22G
do
Highest purity
do.
Trace.
339
do
Chemically pure...
do
Mallinckrodt
14
732
do
B. and C
14
669
do
do
Mallinckrodt
None.
595
do . . . .
.do
do
None.
449
Nitrite
Sticl<:s, C. P
do
E. and A
537
609
do
B. and A
2,130
230
Sulphate
Crystals, reagent. .
Chemically pure...
do
Merck
None.
do
Ceo. D. Feidt
28
895
do
B. and A
21
250
Bisulphate
.do
Crystals, reagent . .
Chemically pure . . .
do
do
Merck
None.
403
Mallinckrodt
Trace.
912
do
do
Geo. D. Feidt
Trace.
1293
B. and A
3
Table 2. — Chlorids in sodium salts.
Serial
No.
Salt.
Description on label.
•
Source.
Parts per million.
Sodium
chlorid.
Chlorin.
268
B. and A..
None.
None.
None.
Trace.
Trace.
187
70
23
46
None.
140
46
None.
None.
None.
46
92
None.
70
46
None.
140
326
326
46
46
70
2,340
56
174
67
71
77
6,143
203
211
None.
None.
None.
None.
117
47
None.
280
163
47
23
None.
None.
23
70
614
do
do .
do
None.
1174
do
1326
do
Chemically pure, crystals.
Chemically pure
B . and L
Trace.
892
Biborate
B. and A
Trace.
1289
1317
Tetraborate
do
Chemically pure, medici-
nal and toilet use.
do
Pacific Coast Borax Co. .
do
113
42
232
Bicarbonate
do
Merck
14
233
Powdered, H. P .
do
28
623
do
Chemically pure
Mallinckrodt
None.
633
do....
do
B . and A
85
235
Carbonate.
Anhydrous, H. P
Dried, reagent
Merck
28
236
do
do
None.
301
do
do
None.
510
. ..do
Aiihydrous, C. P .
B. and A
None.
514
do
do
Mallinckrodt
28
570
do
Anhydrous, H. P
Anhydrous, C. P
Merck
56
655
do
do
B. and A
None.
828
do
do
42
895
do
Anhydrous, H. P
Anhydrous, reagent
Merck
28
956
do
do
None.
1048
do
Anhydrous, C. P
B. and L
84
1128
do
do
.. ..do
198
1135
do
.. ..do..
A. H. Thomas
do
198
1142
... .do... .
Crystals, C. P
28
1146
do
do
do
28
1172
Citrate
U. S. P
do
42
239
Ilydroxid . .
Purified reagent
Merck
1,418
240
do
Pure reagent
do
34
391
do
Mallinckrodt
106
458
do
.. .do
Merck
41
459
..do
Pure reagent
do
43
473
do
Pure by alcohol
Mallinckrodt
Merck
47
551
do
3,723
1419
do
B. and A
do
123
1424
do.
do
128
do
Pure from sodium
Merck
None.
242
Nitrate
do
484
do
do..
Chemically pure
Mallinckrodt
None.
619
do
do
None.
692
do.....
do
B. and A
71
610
Nitrite
.. ..do
do. ..
28
871
do
do
do
None.
1189
Peroxid
Roessler & Hasslacher . .
.do
170
1227
do
99
1383
do
Free from S. P. and CL...
Dried twice, purified
Chemically pure
B. and L
28
244
Phosphate
do. . .
Merck
14
266
Mallinckrodt
None.
409
do
do ^
do
842
do...
.. ..do
B. and A
14
1049
do
do
B.andL
42
186
Table 2. — Chlorids in sodium salts — Continued.
Serial
No.
Salt.
Description.
Source.
Parts per million.
Sodium
chlorid.
Chlorin.
1308
874
Phosphate
Salicylate...
Chemically pure
Crystals, H. P
B. and A
Merck
do
do
B. and L
.... 23
' None.
....j 12
....! None.
None.
14
234
305
1327
Bisulphate
do
do
do
Sulphate
do
do
do
do
do
do
Keagent
Pure, dry reagent
Chemically pure
do
do
do
do
do
do
do
do . ...
None.
None.
1431
267
~ 506
507
552
680
944
1046
.....do
Malliiickrodt
do
do
do
B. and A
Geo. D. Feidt
B. and L
' None.
23
12
12
.... 2,925
70
.... 23
70
None.
14
7
1,772
42
14
42
1212
do
do
do
70
None.
I 70
42
.. ..do
. ..do
do
429
Sulphite.
Crystals
Mackall Bros
42
1444
do
do
A. H. Thomas
B. and A..»
do
:::: 47
....' 140
....! 47
' 93
28
413
do
Hyposulphite
do
Chemically pure, dry
Chemically pure . . .
84
28
841
do
B. and C
54
1305
690
do
do
do
do
B. and A
B. and C
....i 117
47
70
28
1400
Na and K tartrate
Rochellesalt, C. P
Mallmckrodt
1 None.
1
None.
Table 3. — Chlorids in ammoniuin salts.
Source.
Parts per million.
Serial
No.
1
Salt. 1 Description on label.
Ammo-
nium,
chlorid.
Chlorid.
865 ' Acetate
1606 do
189 Carhonste
Chemically pure
do
B.andA ' 4*
Baker i 15
Merck ' 10
3
10
7
832
do
do
None.
do
do
None.
15
42
None.
None.
194
Nitrate
do
do
Crystals
Merck
.do
10
195
28
397
Chemically pure
Mallinckrodt
None.
562
. ..do
do...l
. ..do
None.
None.
^
None.
None.
1123
do
do
do
do
B. and L
None.
1342
do
3
196
Oxalate
do
do
do
do
Nitrate
. ..do
Merck
None.
931
B. and A 21
14
1134
do
. ..do
do.
B. and L
None.
21
30
21
30
21
52
None.
14
1500
.do . .
.do
21
814
■ 678
do
do
B. and A
do
14
21
.do
.do . .
.do
14
197
Phosphate
NH4 and Na phos-
phate.
Sulphate
do
do
do
Reagent
Merck
35
624
Chemically pure
Mallinckrodt 21
Merck 4i
14
198
Reagent
3
• 513
Chemically pure
Mallinckrodt
21
378
14
621
.do... I
. .do
252
664
rin
do.. .
None.
665
do 1 do
do do
do do
Iron alum do
B. and A.
3
829
do
3
.do
3
401
Mallinckrodt
21
187
Table 4. — Chlorids in calcium salts.
Serial
No.
Salt.
Description on label.
Source.
Parts per million.
Calcium
chlorid.
Chlorid.
Carbonate
do
B. and A
10
10
66
42
112
31
31
10
10
21
None.
112
31
1,118
88
7
esi
do
do
7
825
do
do
do
42
5/5
Oxid
.do
.do..
28
(ill
do
do
do
71
1122
.do
..do
do
21
do
.do
.do
21
do
do
do
7
431
Phosphate
do
Diabasic reagent
Merck
7
()3(i
.do
14
1319
do
Diabasic, reagent
do
None.
1387
do
Chemically pure
.do
Baker
71
823
B. and A
21
897
do
do
Mallinckrodt
711
1508
do
do
B and L
56
Table 5. — Chlorids in harium. salts.
Serial
No.
Salt.
Description on label.
Source.
Parts per million.
Barium
chlorid.
Chlorin.
974
Acetate
•
Kahlbaum
63
42
21
21
21
Trace.
126
84
63
42
42
336
None.
42
21
21
21
703
B. and A
14
866
do
do
.do ....
1090
do
do
do
7
1116
..do
. ..do
. ..do
7
1450
do
.do
.do. ..
Trace.
328
Hydroxid
do
Henry Hell
42
462
...do
. ..do
.do
28
612
.do. ..
.do.. .
B. and A
21
947
do
do
A H Thomas
14
1449
. .do
,. ..do
B. and A
14
201
Merck ...
112
471
do
Chemically pure
B. and A
None.
1083
do . .
. . .do
do
14
1117
.do
.do
.do .. ....
7
1425
do
do
do
7
Table 6. — Chlorids in magnesium, salts.
Salt.
Description on label.
Source.
Parts per million.
Serial
No.
Magne-
sium
chlorid.
Chlorid.
1525
Oxid
S. free
B. and L
375
959
141
9
28
19
280
583
do
do
Nitrate
B. and A
710
942
S free
A. H. Thomas
105
558
Chemically pure
Mallinckrodt
7
482
.do.
B. and A
21
488
do
do
do
14
The results given in Table 1 indicate that the occurrence of objectionable amounts
of chlorids in potassium salts is rather the exception, since nearly half of the samples
examined contained no chlorids or only traces. Of the other samples it is seen that the
chromate, nitrate, and hydroxid are the only ones in which the occurrence of excessive
amounts of chlorids appears to be the rule. The acetate and bisulphate samples exam-
ined were found to be particularly free from this impurity. Of the other samples the
amounts found are quite variable, and no conclusion in regard to the source of the
chlorids present can be drawn.
The results included in Table 2 show that the acetate samples are practically free
from chlorids. The borax samples are supposed to be of ordinary commercial quality,
188
and the amounts of clilorids found are no larger than might be expected. The bicar-
bonate and carbonate samples are quite variable, but it is seen that many of the samples
are apparently free from chlorids,. and therefore its elimination from material claimed
to be chemically pure is evidently only a matter of care and experience on the part of
the manufacturer. Passing over the one sample of citrate, it is seen that all the samples
of hydroxid, with the single exception of the material made from metallic sodium, con-
tain amounts of chlorids which vary over very wide limits. It therefore appears that
in all samples of sodium hydroxid, except that made from the metal, the occurrence of
chlorids is to be expected, and an allowable limit of this impurity in the best quality of
goods should be fixed. From the results obtained it would appear that not exceeding
40 parts of chlorid per million should be a safe limit to adopt. The samples of nitrate,
and also of nitrite, are apparently low in chlorids. The samples of peroxid, however,
which are in fact sold as chlorid free, contain amounts which are appreciable, although
probably not beyond the limits which would interfere with the use of the material for
the determinations in which it is employed. The samples of phosphates examined
show small amounts of chlorids, and it appears that such material can readily be pm*-
chased chlorid free. Although the bisulphite samples show small amounts or no
chlorids, the samples of sulphates, sulphites, and hyposulphites are found to contain
small amounts very generally distributed. One sample of sodium sulphite, for some
unaccountable reason, is very high in the amount of this impmity.
The ammonium salts as exhibited in Table 3 are on the whole but slightly contam-
inated with chlorids. The amounts found, except, perhaps, in one case of ammonium
sulphite, are too small to materially affect the quality of the samples.
The calcium and barium salts practically all show small amounts of chlorids, and its
absolute elimination would probably require more effort than the advantage gained
would repay.
In the case of magnesium oxid the presence of excessive amounts of chlorids is of
general occurrence. In the nitrates and sulphates the amount present, however, is
very much less.
The observations in the tables are only of general applicability. The results are
mainly useful in indicating the general occmTence of chlorids in the chemicals at
present found on the market.
The report on taiinin ha\dng been referred to Committee B on
recommendations of referees, after their adjournment ^[r. Kebler
offered a supplementar}-' report on behalf of the committee to the effect
that the work on tannin be continued by the association, and said
recommendation was adopted.
The Presidext. We have with us a visitor distinguished in this
country and abroad, and I am sure that those who liave not met him
will be ver}^ glad to do so. I am going to call on Doctor de 'Sigmond^
professor of agricultural chemistr}^ in the University of Budapest, to
address the association.
Doctor de 'Sigmond spoke in part as follows:
ADDEESS BY DR. ALEXIUS DE 'SIGMOm.
Mr. President and Gentlemen of the Association: I am glad to have an oppor-
tunity of speaking to you of the excellent work of this convention. Especially in
natural sciences — and agiicultural chemistry is one of the main divisions of applied
science^cooperation is much needed. In my researches I have found that while we
may be sure of the exactness of our experimental results, we can not always be sure of
the correctness of the conclusions drawn therefrom, and by cooperation we are able to
189
strengthen the weak parts of our conclusions l)y comparing our final results. This
scientific cooperation should not only be national, but international, and I think I can
perhaps contribute something to your work from mine.
I was particularly impressed by the address of the president of the association, and
the line of work discussed by Doctor Hopkins is one with which I am very familiar,
being in some ways similar to the work of Liebig, in Germany. In fact, the conditions
in this country in some respects are similar to those in Hungary. The cornfields in
Illinois and the wheat lands of the Northwest are like sections of our country where
corn and wheat are the main crops. The same problem discussed by Doctor Hopkins,
in regard to the proper manuring system, arises, and to prove whether the yield will
decrease without fertilization more than 500 farm experiments were conducted, and in
70 per cent of these the phosphoric acid not only produced an effect, but proved
profitable. The question, however, arose as to whether the farm experiments were
conclusive, inasmuch as in several cases the results of the work were vitiated by con-
ditions which could not be controlled. In order to provide a more accurate method of
determining the need of fertilization, the experiment station of plant industry at
Mazyar-0v4r, of which I was until lately the chemist, inaugurated about ten years ago
the so-called Wagner experiments. These we found generally satisfactory, but great
care was necessary, and considerable time — at least a season — inasmuch as we found
evidence that different results were obtained during the fii-st stages of growth and at the
time of harvest. As the farmer wants these results more quickly, we turned to the
chemical laboratory for a solution. We all know that the chemical methods for the
determination of available plant food in the soil are not satisfactory; that is, they do not
always agree with the results obtained in practice. This question I studied for five or
six years, and my results may add something to your cooperative work.
The first question was to determine the most suitable solvent for the determination
of available phosphoric acid. Schlosing found that if he treated a soil with pure water
and gradually increased the amount of nitric acid in the water the quantity of phos-
phoric acid dissolved rapidly increased until a level was reached at which the increase
ceased. The points at which these changes take place vary widely in different soils.
Above these limits the amount of phosphoric acid dissolved again increases rapidly.
This indicates to me that there is a difference between the natural solubilities of the
phosphates in the soil which divides the phosphoric acid present into the soluble and
the less soluble. I studied twelve different soils which all corroborated this idea.
The next point was to find out whether there is any relation between the quantity
of slightly soluble phosphoric acid present and the amount needed in the soil.
Experiments were made with 100 different soils analyzed by my method and tested
both by field and pot experiments, and I was able to figure out in this way the practical
limits of the phosphoric acid needed in the soils. At the same time I found that the
reaction of the soil must be taken into consideration in drawing conclusions from the
experimental results. For example, when an almost neutral soil needed phosphate,
the slightly soluble phosphoric acid present never exceeded 0.007 to 0.015 per cent.
On the other hand, in soils on which the phosphates produced .practically no effect
there was at least 0.075 per cent or more of soluble phosphoric acid present. This max-
imum limit was corroborated in all cases, but for the minimum limit — that is, the one
showing the need of phosphoric acid in the soil — further classification was necessary.
For example, in several cases soils rich in calcium carbonate and containing almost
0.050 per cent of soluble phosphoric acid still showed in the fertilizing experiments
the need of phosphates. According to the experimental results, I was obliged, there-
fore, to classify the soils according to their basicity, it being the general rule that the
increase of basicity of the soil decreases the availability of the soluble phosphates.
We also made extensive studies of our alkali soils, and it was gratifying to me when
in California to find that the results obtained by Professor Hilgard are in close relation
to those which I have found for such soils. I might state here that I have also used
190
with satisfaction some of the practical surveying methods of the Bureau of Soils of the
United States Department of Agriculture.
Another question studied, which is not new but needs further attention, is the need
of special fertilizers by different plants. We have found, for instance, that wheat
needs a great deal of available phosphoric acid in the early period of its deA^elopment,
while com can take it later; therefore, when we fertilize corn, a phosphate of lower
availability can be used than for wheat, because the corn can utilize a less available
form.
I will not pursue this discussion further, but will conclude my remarks with a cordial
invitation to each member of this association to come to our country and investigate
our experimental fields and work; you will find us investigating problems very similar
to your own.
After some discussion the following motion was passed as to the
place of meeting for the convention of 1907:
Resolved, That the place of meeting in 1907 be left to the discretion of the executive
committee, the preference of the association being understood to be for Xorfolk, if
deemed practicable, when the call for the meeting is issued.
Mr. Veitch. Last year the referee submitted a revised method for
the determination of tannin. The committee on revision of methods
has made some changes, chiefly in manner of statement, etc., in this
method, and I move that the method as revised by the committee
be substituted as the provisional method of the association.
The motion was carried.
EEPOET OIsF POTASH.
By A. L. KxiSELY, Referee.
The work this year has been dei'oted to the determination of potash in one sample
of soil and in one sample of mixed fertilizer containing considerable organic matter.
Upon the sample of soil the official and the proposed volumetric methods were used ;
upon the sample of fertilizer the official methods for water soluble and for total potash
(K2O) were used; also the proposed ignition method, the proposed volumetric method,
and Carpenter's method.
The directions sent to the cooperating chemists Avere as follows:
AssociATiox Potash Work.
Sample No. 1. — A t\'pical soil from the wheat -growing belt of Sherman County. Oreg.
Ninety-nine and seven-tenths per cent of this soil passes readily thi'ough a 0.5 mm
sieve. This soil contains, approximately. 1.50 per cent of total potash (J. Lawrence
Smith method) and from 0.25 to 0.5 per cent acid soluble potash.
Sample Xo. 2. — A mixed fertilizer containing considerable nitrogen, phosphoric
acid, and potash — approximately 7 to 9 per cent of potash (KoO).
ASSOCIATIOX WORK UPOX SOIL.
Mix thoroughly before sampling.
1. Determine potash (ICO) in soil according to official method using hydrochloric
acid 1.115 sp. gr. (Bui. 46, revised, p. 71).
2. Proposed volumetric method for potash in soils.
191
Reagents.
Nitric acid — 5.5 cc nitric acid, 1.40 sp. gr., in 1,000 cc water.
Sodium nitrate wash — 5 grains sodium nitrate per 1,000 cc water.
Phosphomolybdic solution— 100 grams phospliomolybdic acid (Kahlbaum's
preferred) in 750 cc water and 250 cc nitric acid 1.40 sp. gr.
Standard solutions— Standard caustic potash and nitric acid prepared for volu-
metric phosphoric acid. One cc of potassium hydroxid is equal to 1.655 mg of
potash (KoO).
Determination.
Take a fresh aliquot portion, representing 1 gram, of solution A under soil analysis
(Bui. 46, revised, p. 72) in a porcelain dish, add 10 cc phosphomolybdic solution and
evaporate to dryness. Add 25 cc nitric acid wash heated to 50° C, and stir thor-
oughly. Allow to cool, filter through a thick asbestos filter, and wash with sodium
nitrate wash until free from acid; transfer the gooch to a tall beaker, add an excess of
standard potassium hydroxid and heat nearly to boiling (any precipitate adhering to
the crucible should be dissolved with the standard alkali). When the precipitate is
completely dissolved add a few drops of phenolphthalein, which should show an excess
of alkali, and titrate back with standard nitric acid.
ASSOCIATION WORK UPON MIXED FERTILIZER.
Mix thoroughly before sampling.
1. In sample No. 2 determine water soluble potash (KoO) according to the official
method (Bui. 46 revised, p. 21).
2. In sample No. 2 determine total potash (K2O) in organic compounds by the
official method (Bui. 46, revised, p. 22, (6) With organic compounds).
3. Proposed method. — Ignite 5 grams of sample No. 2 at dull redness to a gray ash.
Transfer to 200 cc beaker, using 25 cc of hot approximately 10 per cent hydrochloric
acid to remove last traces, cover with watch glass and let simmer on hot plate for
one-half hour; then add 100 cc water and to hot solution add slight excess ammonia
and ammonium oxalate, cool, make up to 250, filter through dry filter. Evaporate
50 cc of filtrate and proceed as under 3 (a) in mixed fertilizers (Bui. 46, revised, p. 22).
4. Proposed volumetric method. — Ignite 5 grams of sample No. 2 at dull redness to a
gray ash. Transfer to 200 cc beaker, using 25 cc of hot approximately 10 per cent
hydrochloric acid to remove last traces, cover with watch glass and let simmer on hot
plate for one-half hour, cool, make up to 250 cc, filter through dry filter, take 25 cc of
filtrate in a porcelain dish, add 20 cc phosphomolybdic solution, and evaporate to
dryness; add 50 cc nitric acid wash heated to 50° C. and stir thoroughly. Proceed
as under proposed volumetric method for potash in soils.
5. Carpenter' s proposed method. — Boil 10 grams of the sample with 300 cc of water
plus 5 cc of hydrochloric acid for thirty minutes. Add a few drops of phenolphthalein
and carefully neutralize with sodium hydrate free from potash, avoiding a large
excess. Add sufficient powdered ammonium oxalate to precipitate all the lime
present, cool, dilute to 500 cc, mix and pass through a dry filter. Complete deter-
mination by official method (Bui. 46, p. 22).
In each case run blanks to ascertain corrections to be made for impurities. (It is
necessary to ascertain blank in phosphomolybdic solution.) It is also advisable to
treat the potassium platinic chlorid residue in the gooch crucible in order to ascertain
if it is all soluble in water. Then reweigh the crucible after thoroughly drying.
If any workers have time, it is suggested that they make a determination of potash
in each of the two samples according to a method suggested by F. P. Veitch (J. Amer.
Chem. Soc, January, 1905, pp. 56-61).
A. L. Knisely, Referee.
B. B. Ross, Associate Referee.
192
Comj^arUon of official and modified methods for the determination of 'potash in soil and in
mixed fertilizer .
Soil.
Fertihzer.
Analyst.
Official.
Volu-
metric.
Official
water-
soluble.
1 Pro-
Official posed
total. 1 ignition
1 method.
Pro-
posed
volu-
metric.
Carpen-
ter's
method.
-* !
Stillwell and Gladding, Xew|
York City 1
Per cent.
Per cent.
"""6." 48""
Per cent.
8.50
8.77
-i 8.71
8.64
8.76
8.77
8.71
Per cent.
Per cent.
9.40
Per cent.
■Per cent.
9.00
W. D. Richardson, Swift & Co.,
8.93
8.80
8.76
8.81
8.57
8.62
8.74
C7.96
C7.99
8.98
9.00
Chicago
8.96
R. W. Thatcher, Washington
station .
{ .40
.39
1
a. 38
6.40
a. 41
6.40
.45
.47
.51
C7.95
C7.97
8.95
8.98
8.85
C7.99
C7.89
W. E. Dioldnson, South Caro-
r .47
I .45
.47
.47
.41
.41
1
1
lina station
1
i
B. F. Robertson, South Caro-
8.73
8.73
9.04
9.00
8.85
8.74
9.13
9.05
8.83
8.81
8.73
8.74
8.79
8.69
9.08
9.12
9.14
8.98
8.95
J. H. Mitchell, South Carolina \
.40
.43
.39
8.92
L. Heimburger, Agricultural
F. A. Welton, Ohio station.
.48
1 -
8.82
8.67
8.72
8.89
9.09
8.84
8.82
M. G. Donk, Bureau of Chem-
istry, Washington, D. C
■1
.46
.39
.41
.40
8.84
9.45
9.60
9.27
8.73
8.82
9.25
9.04
C. E. Bradley and A. L. Knise-
■-.■41""
.42
.42
.41
.44
8.49
8.49
8.50
8.82
8.83
8.17
8.24
8.36
8.89
8.99
Average
.439
.426
8.71
8.89
8.78
8.97
8.98
Maximum
Minimum
.48
.39
.51
.38
9.04
8.49
9.13
8.74
9.40
8.17
9.60
8.57
9.25
8.82
Difference
.09
.13
.55
.39
1.23
1.03
.43
a Without evaporating to dehydrate silica.
6 After evaporating to dehydrate silica.
c Omitted from average.
Since the foregoing results A"aiy considerably, the question may be raised as to the
proper sampling of the fertilizer. Samples were prepared as follows:
The mixed fertilizer was finely ground and sifted upon a large piece of flexible oil-
cloth. The sample was thoroughly mixed by rolling alternate corners of the oilcloth.
After thorough mixing, about 10 grams of the sample were placed in each of 40 sample
bottles, the fertilizer remaining on the oilcloth was again thoroughly mixed and a sec-
ond 10 grams put in each sample bottle. This process of mixing and sampling was
continued until all the fertilizer was used.
The contents of each sample bottle of fertilizer was mixed by shaking and rolling,
after which the contents of all sample bottles were again emptied on the oilcloth and
thoroughly remixed. The sample bottles were again filled in the manner previously
described and immediately sealed with paraffin.
Com:mexts by Analysts.
Thomas S. Gladding: The results obtained are interesting and seem to show that the
method by ignition and solution in acid is the only method that gives total available
potash present.
R. JV. Thatcher: The volumetric method seems to me to be very promising. I find
difficulty in getting anj-thing like a satisfactory sample of phosphomolybdic acid from
any of our American dealers, and think that it will be practically necessary for each
analyst to prepare his own material for use in this method.
193
W. E. Dickinson: It seems to me that the volumetric method is too greatly affected
by varying conditions to be used successfully by beginners.
L. Ileimburger: In the volumetric method, with both samples No. 1 and No. 2, the
best results were obtained with smaller samples than recommended, due probably to
the difficulty encountered in washing the potassium phosphomolybdatc free from all
the excess of phosphomolybdic acid.
F. A. Welton: Concordant results were not obtained by the volumetric method on
either sample.
M. G. Donk: I was unable to get concordant results using 0.5 gram of material, as
directed, as the precipitate on using this quantity of material was of such a coherent
nature that it was not possible to wash it entirely free of the excess of reagents used.
The results reported were obtained by using 0.1 gram of material, 5 cc of phospho-
molybdic solution, and 20 cc of nitric acid wash.
Comments by Referee.
Nitric acid heated to 50 ° C. exerts a marked solvent action on the potassium phos-
phomolybdatc. This is especially marked in soil work where potash is present in
small amount. For this reason cold nitric acid saturated with pure potassium phos-
phomolybdatc was used instead. The acid solution of the soil was directly evapo-
rated with the phosphomolybdic acid, as results by this method agreed closely with
those made on solution A, Bulletin 46, page 72.
No difficulty was experienced in making results on the fertilizer by the official
method agree when made on the same solution; but different solutions of the same
sample, supposed to be made in identically the same way, gSve results that in some
cases varied considerably.
The method of ignition and subsequent solution in acid gives slightly higher results
than the official method for water soluble potash. The referee does not believe that
this extra potash should be placed on the same basis with the water soluble.
Additional Results.
Some additional results, received too late to be included in the averages, are given
in the following table :
Additional potash determinations made on referee'' s sainples.
•
Soil.
Fertilizer.
Analyst.
Official
method.
Official
water
soluble
method.
Proposed
ignition
method.
Proposed
volumet-
ric
method.
Corpen-
ter's pro-
posed
method.
G. S. Fraps, Texas station ^
Per cent.
0.63
Per cent.
{ 9.01
\ 9.06
[ 9.02
Per cent.
Per cent.
f 8.99
1 8. 98
Per cent.
8.90
8 96
8.88
8.76
9.12
8.99
9.03
9.04
8.97
F. G. Keves, Rhode Island station
9.03
9.05
9.05
W. F. Purrington, Rhode Island station
Mr. Hartwell writes as follows in regard to the work reported from the Rhode Island
station :
In the case of the gravimetric method, Mr. Keyes took 5 grams of the sample and
made it up to 250 cc, taking only 25 cc for each determination, whereas Mr. Purring-
ton made the solutions in a similar manner, but took 50 cc for each determination,
or the equal of a half gram of the sample, as is required by the official method.
31104— No. 105—07 18
194
The blank determination with the gravimetric method vrsis equal to 3 mg of potas-
sium chlorplatinate, and this blank has been deducted from each analyst's results
before reporting the percentages. In the case of the volumetric method, the blank
^vas equal to 0.2 cc of the standard potassium hydroxid solution. This blank has
likewise been deducted before reporting the results, and in all cases a correction was
made for calibration of the glassware. In the case of the gravimetric results, the fac-
tor 0.194a was used to convert the potassium chlorplatinate to potassium oxid.
Recommexdatiox.
The referee recommends a continuation^of the study of the volumetric method for
use in both soil and fertilizer analysis.
^Ir. CusHMAX. It seems to me that the association should consider
this important question of potash determination in another way.
The work consists now in a comparison of methods for the deter-
mination of total and water-soluble potash^ but should we not
determine whether there is any potash present and available that
is not water soluble ? It is possible to put a definite weight of potas-
sium chlorid into an acid-phosphate fertilizer and then have the
official analysis show less potash than was added. Again, it is possible
to make up a solution of potassium chlorid, and after it has been
shaken with a natural clay some of the potash disappears fi^om the
solution into the clay, but it is not possible to wash the clay and
regain the potash. The point I am making is that the method of
leaching with water does not tell the truth, for the potash held by
absorption in the clay would be available as plant food. Again, in
a wet feldspar powder three kinds of potash must be recognized —
that in the crystalline compound which has not yet been set fi^ee,
that set free but held by absorption in the decomposed product, and
that which goes freely into solution in water. Is it true, then, as so
many chemists believe to-day, that only one of these three kinds of
potash is available as plant food ? My belief is that two of these are
available, the water soluble and that held by absorption in the
colloidal decomposition products of the material. I have extracted
from a feldspar containing 9 per cent of total potash 3 per cent,
using only water, but that same sample referred to a commercial
chemist and analyzed by the official method was retitrned as con-
taining only 0.1 per cent of available potash. I maintain that ottr
methods are unsatisfactory as long as sttch a condition is possible.
If I can extract from a finely-ground feldspar by the use of distilled
water a large part of the potash present, it is possible that the same
thing takes place in nature, and there is no e^-idence to show that by
leaching a finely-ground material ^\'ith water all of the potash avail-
able as plant food is obtained. On the other hand, there is abundance
of evidence to show that there is more potash present available as
plant food that does not go mto water solution under the conditions
a This factor is given by Konig in Die Untersuchung landwirtshaftiich und gewer-
blich wichtiger Stoffe, 1906, p. 30.
195
of our present analytical process. In view of these facts, I wish to
recommend that the referee direct his attention to a study of what
really constitutes available potash in soils, fertilizers, and ground
mineral products, so that, if possible, we may have a definition of
available potash. The refcEee, I understand, according to the pre-
cedents and rules of our association, has the right to call in subrcferees,
or at least other members of the association, to aid him in making
his report. If there were an}^ machinery for doing it, I think the
matter so important that I would suggest the appointment of a
representative committee to study the question carefully, but as it
is I recommend that the referee turn his attention to a definition of
available potash.
Mr. Hart WELL. It seems to me that Mr. Cushman is laboring
under a misapprehension as to the attitude of the association.
Hardly any one experienced in soil work would claim that the water-
soluble potash necessarily represented the available potash in a given
fertilizer. We have, however, in many of the States, laws which
distinctly state that potash shall be determined as water soluble.
This is not saying that it represents the available potash; in fact,
if we are using the plant, the only true measure of the availability,
as the standard, the availability will vary with the plant used. What,
then, shall be considered as available potash ? That which is available
to the turnip or to some other crop ? These questions are well recog-
nized by the members of the association and will, I think, be con-
sidered in future work. Terhaps it might be well for the referee
to conduct pot experiments for the purpose of determining this
question.
^h\ Frear. The purpose of analyzing fertilizers is primarily to
determine to what extent they contribute plant food to the soil, and
while the general principles are fairly well understood they are subject
to change as the work progresses. I think we have all been deeply
interested in the results of Mr. Cushman's work with minerals which
are ordinarily supposed not to give up their basic substances within
any short interval. But the problem which confronts the chemist
is this : Here is one class of materials conceded to contain immediately
available plant food and another class which contains the same
ingredients but in a form much less quickly available if at all so.
The temptation is to resort to the use of the latter class, at the same
time guaranteeing the same high percentage of nutrient constituents.
While critical studies along this line are in place, I think the work of
the chemist must for some time continue to consist in the distinction
between materials of high and low value.
^'Ir. BowKER. I do not think that there is a fertilizer manufacturer
of any repute who desires to introduce into his goods insoluble forms
of potash, for if the analytical tests were not satisfactory, the results
196
in the field would not be. and after all the final test is the field
test. Do they give satisfaction to the farmer '? TTe are using large
quantities of cotton-seed meal and other legitimate substances which
contain organic forms of potash. Under the present methods of
examination no credit is given for that potash, and I think the
methods ought to be changed to correct this. TThen we make an
acid phosphate and add to it muriate of potash, about 10 per cent of
phosphoric acid and S percent of potash, there is. for some unaccoimt-
able reason, a loss of fi'om 5 to 10 per cent of the potash present accord-
ing to the official retm-ns. I ask this association, as the officials
who stand between us and the consumer, if that is fab? It is true
the laws say that water-soluble potash shall be deternnned. but those
laws are based. I think, on the Massachusetts law. which was framed
back in 1S73. before we had experiment stations, when the idea
was that only water-soluble materials were available. The original
laws called for water-sohible phosphoric acid, but to-day reverted
phosphoric acid is recognized and determined by an arbitrary method.
I maintain that if it is right to recognize two forms of phosphoric
acid there should be methods for deternnning two forms of potash.
We may ask for an amendment to the law in Massachusetts this
vear which will provide for such deterroinations. and I think our
experiment station will work with us. In my opinion ^Ir. Cushman
Ls doing a gi'eat work. Every year we pay enormous sums to Germany
for imported potash, but the experiments performed by !Mr. Cushman
before the chemists m Boston in obtaining potash from finely groim^d
rock by electrolysis open tip the possibility of obtaining om- potash
from the feldspar deposits in this coimtiy.
^Ir. Cushman's motion was referred to conunittee A for action.
EEPOET OF COMMITTEE A 01s EEGOMMENDATIONS OF EEFEEEES.
By E. J. Davedsox. Chairman.
(1) XlTKOGEX.
It is recommended:
1. That the work on the Fuller modification of the official Gunning method be dis-
continued, and that the Gunning method remain as it is now stated.
Adopted.
2. (a) That the work on the neutral permanganate method be continued along the
same lines as this year, having in mind the influence of excessive amounts of nonnitrog-
enous material in the source of nitrogen, (b) That work be directed along the line of
eliminatiug some of the many detaib of the method which influence to tod great a
degree the results obtained.
Adopted.
3. That the work on the alkaline permanganate method be continued and that the
quantity of material taken be changed to 0.0675 gram nitrogen: that the quantirj-
of alkaline permanganate used in digestion be changed to 150 cc. and that 100 cc be
distilled oS before titration. The modified method should read as given on page S3.
Referred to the referee for 1907 for iavestisration as to whether this method or some
197
modification of it can be applied to all organic nitrogenous materials, especially cotton-
seed meal.
4. That the method for standardizing hydrochloric acid proposed in 1905 and
referred to the referee for 1906 be adopted and that the method now given in the.
official methods be dropped. [See Bui. 46, Rev., p. 14, under "4. Determination
of nitrogen." The proposed method reads as follows:
By means of a preliminary test with silver-nitrate solution, to be measured from a
burette, with excess of calcium carbonate to neutralize free acid and potassium chro-
mate as indicator, determine exactly the amount of nitrate required to precipitate all
the hydrochloric acid. To a measured and also weighed portion of the standard acid
add from a burette one drop more of silver-nitrate solution than is required to precipi-
tate the hydrochloric acid. Heat to boiling, cover from the light, and allow to stand
until the precipitate is granular. Then wash with hot water through a Gooch cruci]:>le,
testing the filtrate to prove excess of silver nitrate. Dry the silver chlorid at 140° to
150° C.
Adopted as the official method.
[Note by the secretary. — The following method for the estimation of nitrogen
was submitted by Mr. T. S. Gladding, together with comparative results on 81 samples
obtained by the three methods compared, the question being raised as to whether a
combination of two official methods, the Gunning and the Kjeldahl, would be consid-
ered an official method. Mr. Gladding requested that appropriate action be taken,
and the method is submitted for the information of the referee without instructions ■
from the association:
COMBINATION METHOD: KJELDAHL AND GUNNING METHODS.
One gram of fertilizer; 25 cc of sulphuric acid; 10 grams of potassium sulphate; 0.7
gram of mercuric oxid; heat till water white. Cool, add 200 cc of water, 0.5 gram of
zinc dust, 25 cc of potassium sulphid solution, 50 cc of soda solution, and distil.]
(2) Separation of Nitrogenous Bodies.
A. milk and CHEESE PROTEIDS.
It is recommended:
1. That the original method of preparing the water extract [cheese analysis], pro-
posed by Van Slyke and Hart, be so altered as to call for the use of 1,000 cc of water
in place of 500 cc. [Proceedings, 1902, Bui. 73, p. 89.]
Adopted.
2. That the method of drawing the water extract through a thick pad of asbestos,
after it has been separated from the fat and insoluble nitrogenous matter by cotton
wool, be further studied.
Adopted.
3. That the temperature at which the extraction is made be further studied.
Adopted.
4. That the completeness of the extraction of matter soluble in salt solution be
further studied.
Adopted.
B. VEGETABLE PROTEIDS.
The report of the referee was received too late for action to be taken by committee
A, but the following recommendation is submitted as a matter of record:
It is recommended, for the purpose of securing greater uniformity of results, that
in the extraction of alcohol-soluble nitrogen in wheat and ffour 70 per cent alcohol by
weight, sp. gr. 0.871, be used.
198
C. iEEAT PEOTEEDS.
It is recommended:
1. Tliat the modified Tannin-salt method as described in the proceedings of the asso-
ciation for 1905 [Biireati of Chemistr\- Bill. Xo. 99. p. 182] be adopted as a provisional
method.
Referred again to the referee for recommendation in 1907.
2. That the xanthin base method of Schiuenhelm [Proceedings. 1904. Chemistry
Bui. Xo. 90. p. 129] be adopted as a provisional method.
Referred again to the referee for recommendation in 1907.
3.^ That the application of the kreatinin method as applied by Folin to the urine be
further studied by the association. [Zts. physiol. Chem.. 1886. 10: 391.]
Adopted.
(3) IXOEGAXIC PlAXT CoXSTlTL EXTS.
It is recommended:
That the peroxid method for total sulphur be adopted as^official.
[This method -w-as recommended as provisional by the referee in 1905. Btil. No. 99.
p. 133. the restatement of the method given on page 152 vatying only in technique.]
Referred to referee for 1907 for recommendation as to final action.
2. That the combustion method li. e.. the Sauer-Tollens-Barlow method ) for deter-
mining volatile inorganic plant constituents be further investigated. [Sauer. Zts. anal.
Chem.. 1873. 12: 32: Barlow-ToUens. J. Amer. Chem. Soc, 1904. 26: 341.]
(4> Potash.
It is recommended:
1. That the study of the volumetric method for use in both soil and fertilizer analysis
be continued. [Proceedings. 1905. Bureau of Chemistry. Bui. Xo. 99. p. 135.]
Adopted.
2. The following recommendation, made by ^Ir. Cushman. was referred to committee
A after their report had been made. The committee, however, considered the recom-
mendation subsequent to the adjournment and referred it to the referee on potash
for investigation.
I recommend that the referee direct his attention to a study of what really consti-
tutes available potash in soils, fertilizers, and ground-mineral products, so that, if
possible, we may have a definition of available potash. The referee. I understand,
according to the precedents and ndes of our association, has the right to call in sub-
referees, or at least other members of the association, to aid him in making his report.
If there were any machinery- for doing it in the association. I think the matter so impor-
tant that I would suggest the appointment of a representative committee to study the
question carefully: but as it is I put it in the form of a recommendation, that the referee
turn his attention to a definition of available potash.
(5) Soils.
It is recommended:
1. That the fifth-normal nitric-acid digestion method be fiuirher studied, [Bui.
Xo. 46. p. 74 (i ). usins: nitric instead of hydrochloric acid and digesting at 20° C. instead
of40°C.] ^
Adopted.
2. That the sodium-peroxid fusion method for total phosphorus be given a further
trial, and that this be compared with the alkali-carbonate fusion method. [Proceed-
ings. 1905. Bui. 99. p. Ill: details sHghtly modified in 1906 report.]
Adopted.
3. That the modified J. Lawrence Smith method for total potassium, presented at
this meeting (p. 147 i, be further tested.
Adopted.
199
4. That line 30, under "1. Preparation of sample," page 71, Bulletin No. 46, be
changed from "openings -^- millimeter in diameter," to "openings 1 millimeter in
diameter," and that "passed through a sieve of 1-millimeter mesh " l)e omitted from
line 1 under "(h), page 74."
This change in an official method, having been before the association in 1905, was
adopted.
5. That for "3. Determination of volatile matter," Bulletin No. 46, page 72, sub-
stitute the Determination of Total Organic Carbon [J. Amer. Chem. Soc, 1904, 26:
1640] as the official method.
Recommended for adoption as official in 1907.
6. That under (k), page 75, Bulletin No. 46, mark the official method "(a)" and
insert the following:
B. OPTIONAL PROVISIONAL METHOD.
Proceed as in (a) through "let stand a few minutes in the water bath" and complete
as follows:
Filter into a beaker, add a drop or two of hydrochloric acid and 1 cc of ammonium
sulphate (75 grams to 1 liter), digest several hours on water bath, and filter into a tared
platinum dish. Evaporate to complete dryness, heat to dull redness, add 1 gram of
powdered ammonium carbonate, expel by heating, cool, and weigh the sulphates of
sodium and potassium. Determine potassium in the usual manner.
Adopted.
(6) Insecticides.
It is recommended that the methods submitted for investigation in 1905 be further
studied. These methods are as follows:
1. Work on London purple, special attention being given to the modifications pro-
posed by Mr. Davidson for the removal of part of the color.
Davidson's modification.
Total arsenious oxid.
Place 2 grams of London purple in a beaker and dissolve in about 80 cc of water
and 20 cc of concentrated hydrochloric acid at a temperature of 80°. Cool and add
sodium carbonate in slight excess, transfer to a 250 cc flask, and bring to the mark;
shake and filter through a dry filter into a dry beaker; acidify 50 cc with hydrochloric
acid and add sodium bicarbonate, titrating with iodiii as usual.
Total arsenic oxid.
Acidify 50 cc of the alkaline solution, prepared as described under "Total arsenious
oxid," with hydrochloric acid, add 25 cc of concentrated hydrochloric acid and 3
gram.s of potassium iodid. Then proceed as directed under this determination in
Circular 10, revised. Bureau of Chemistry, page 4.
2. The hydrogen peroxid method for determining sulphur in sulphur dips and
similar compounds, effort being made to have all analyses made simultaneously.
3. The hydrogen peroxid method for determining formaldehyde in its modified form.
4. The methods of determining available chlorin in bleaching powder.
Motion carried.
EEPOET OP COMMITTEE m THE PEESIDEl^T'S ADDRESS.
The committee appointed to consider the President's address begs leave to submit
the following report:
(1) In view of the general interest of the subject under discussion to others than
agricultural chemists we recommend that a special edition of the president's address
be published separately from the Proceedings for wide distribution.
200
(2) Inasiniicli as there is not sufficient time diu'ing this conA'ention to give the matter
the full consideration which it deserves,, we recommend that a committee be appointed
which shall, after consultation with the Secret-ary of Agriculture, consider in detail
the questions raised in the address and report at the next meeting of the association.
F^^\. WOLL.
L. L. Van Slyke.
A. L. "WiNTON.O
B. B. Ross.
R. J. Davidson,
A. M. Peter.
C. L. Pexxy.
The report of the committee was adopted by the association.
EEPOET OF COMMITTEE ON UI^IPIOATIO]^ OP TEEMS POE EEPOETINQ
ANALYTICAL EESULTS.
]Mr. K. J, Davidson, as chairman of this committee, reported cer-
tain recommendations to the association, statmg, however, that only
meager responses had been received from the chemists considted
during the two years that the matter had been under consideration,
during which time two prehminary reports had been submitted in
circular form for criticism. In the discussion which followed,
participated in by Messrs. Frear, Davidson, Pemiy, Bigelow, and
Hopkins, it appeared that the association was not ready to vote on the
recommendations, and that a wider expression of opinion was desir-
able, especially from the American Chemical Society and the agricul-
tural colleges and experiment stations, before taking any definite
action. Mr. Davidson again called attention to the fact that every
endeavor had been made to obtain an expression of the views of
chemists at large with but slight response, and the committee was
rather at a loss as to how to proceed further in the matter. After
some further discussion, especially as to the merits of the element
system of nomenclature, it was ordered that action on the report be
deferred until another year and the committee continued,
EEPOET OP THE COMMITTEE ON EESOLUTIONS.
By L. L. Tax Slyke.
Resolved, That we extend to the Secretary of Agi'iculture and the Assistant Secre-
tary of Agi'iculttu-e otu' hearty thanks for the continued assistance which they have
so generously given to this association.
Resolved, That we acknowledge the comtesies of the George AVashington University
and thank it for the use of these rooms for our convention.
Resolved, That the secretary convey to the Cosmo? Club an expression of the hearty
appreciation of the members of this association for the cotutesies extended to them
by the chib.
The report of the committee was adopted and the association
adjourned.
a Mr. Winton later resigned, and the vacancy was filled by the appointment of
Ml'. J. G. Lipman.
OFFICERS, REFEREES, AND COMMITTEES OF THE ASSOCIATION
OF OFFICIAL AGRICULTURAL CHEMISTS FOR THE YEAR
1907.
President.
Mr. John P. Street, New Haven, Conn.
Vice-President .
Mr. Harry Snyder, St. Anthony Park, Minn.
Secretary.
' Mr. H. W. Wiley, Washington, D. 0.
Additional members of Executive Committee.
Mr. B. B. Ross, Auburn, Ala.
Mr. B. L. Hartwell, Kingston, R. I.
Referees.
Phosphoric acid: B. W. Kilgore, Raleigh, N. C.
Nitrogen:
Determination of nitrogen: C. L. Penny, Newark, Del.
Separation of nitrogenous bodies: L. L. Van Slyke, Geneva, N. Y. (milk and
cheese proteids).
Potash: A. L. Knisely, Corvallis, Oreg.
Soils: J. H. Pettit, Urbana, 111.
Dairy products: F. W. Woll, Madison, Wis.
Foods and feeding stuffs: J. K. Haywood, Washington, D. C.
Food adulteration: A. E. Leach, Boston, Mass.
Sugar: C. A. Browne, jr., Washington, D. C. (Special analytical methods.)
Tannin: F. P. Veitch, Washington, D. C.
Insecticides: R. J. Davidson, Blacksburg, Va.
Inorganic plant constituents: AV. W. Skinner, Washington, D. C.
Medicinal plants and drugs: L. F. Kebler, Washington, D. C.
Associate referees.
Phosphoric acid: J. M. McCandless, Atlanta, Ga.
Nitroge7i:
Determination of nitrogen: G. W. Cavanaugh,. Ithaca, N. Y.
Separation of nitrogenous bodies —
Meat proteids: F. C. Cook, Washington, D. ('.
Vegetable proteids: Harry Snyder, St. Paul, Minn.
Potash: B. B. Ross, Auburn, Ala.
Soils: S. D. Averitt, Lexington, Ky.
Dairy products: J. M. Bartlett, Orono, Me.
Foods and feeding stuffs: John P. Street, New Haven, Conn.
(201)
202
Food adulteration:
(1) Colors: E. F. Ladd, Agiicultui-al College. X. Dak.
(2) Saccliaxme products, including confectionery- : C. H. Jones. Burlington. Vt.
(3) Fruit products: H. C. L^thgoe. Boston. Mass.
(4) TVine: Julius Hortvet. St. Paul. Minn.
(5) Beer: H. E. Barnard. Indianapolis. Ind. (General associate on food
adulteration, "i
(6) Distilled liquors: L. M. Tolnian. Washington. D. C.
(7) Yinegai-: Charles H. Hickey. Boston. Mass.
(8) Flavoring extracts: E. M. Chace. Washington. D. C.
(9^ Spices: A. L. Winton. Chicago. 111.
(10) Baking powder and baking chemicals: W. M. Allen. Raleigh. X. C.
(11) Meat and fish: E. L. Redfern. Lincoln. Xebr.
(12") Fats and oils: L. M. ToLman. Washington. D. C.
(13) Dain^ products: A. E. Leach. Boston. Mass.
(14) Cereal products: A. McGQl. Ottawa. Canada.
(15) Vegetables: W. L. Dubois. Washington. D. C. '
(16) Condiments other than spices: R. E. Doolittle. Xew York.
(17) Cocoa and cocoa products: E. M. Bailey, Xew Haven. Conn.
(18) Tea and coffee: CD. Howai-d. Concord. X. H.
(19) Preser^-atives: W. D. Bigelow. Washington, D. C.
(20) Determination of water in foods: F. C. Weber. Washington. D. C.
Sugar:
Molasses methods: J. E. Halligan. Baton Rouge, La.
Chemical methods: Fritz Zirban. Audtibon Pai-k. La.
Taimin: M. S. McDowell. State College. Pa.
Insecticides: F. S. Shiver. Clemson College. S. C.
Liorgaiiic plant coiistituents: John W. Ames. Wooster. Ohio.
Medicinal plants and drugs: Charles H. La Wall. Philadelphia. Pa.
SPECIAL COMMITTEES.
Food Staridards.
!Mr. William Frear. State CoUege. Pa., chaii-man.
Ml-. H. W. Wiley. Washington. D. C.
^Ii-. H. A. Weber. Columbus. Ohio.
'Sir. M. A. Scovell. Lexington. Ky.
'Sh. E. H. Jenkins. Xew Haven. Conn.
Fertilizer Legislation.
Mr. H. W. Wiley. Washington. D. C. chairman.
Mr. B. W. Ellgore. Raleigh. X. C.
Mr. H. B. McDonnell. College Park. Md.
Ml-. J. L. Hills. Burlington. Yt.
:Mr. B. B. Ross. Aubiu-n. Ala.
Testing Chemical Reagents.
Mr. L. F. Kebler. Washington. D. C. chairman.
Ml-. A. L. Winton. Chicago. 111.
Ml-. B. W. Kilgore. Raleigh. X. C.
203
Unification of Terms for Re/porting Analytical Results.
Mr. R. J. Davidson, Blacksbiirg, Va., chairman.
Mr. C. G. Hopkins, Urbana, 111.
Mr. W. D. Bigelow, Washington, D. C.
Mr. G. S. Fraps, College Station, Tex.
Mr. C. A. Browne, jr., Washington, D. C.
Committee on the Presidents Address.
Mr. F. W. Woll, Madison, Wis., chairman.
Mr. R. J. Davidson, Blacksburg, Va.
Mr. C. L. Penny, Newark, Del.
Mr. A. M. Peter, Lexington, Ky.
Mr. B. B. Ross, Auburn, Ala.
Mr. L. L. Van Slyke, Geneva, N. Y.
Mr. J. G. Lipman, New Brunswick, N. J.
CoTnmittee on Revision of Methods.
Mr. J. K. Haywood, Washington, D. C, chairman.
Mr. F. P. Veitch, Washington, D. C.
Mr. L. M. Tolman, Washington, D. C.
Mr. J. P. Street, New Haven, Conn.
Mr. A, L. Winton, Chicago. III.
Mr. J. H. Pettit, Urbana, 111.
Mr. F. W. Woll, Madison, Wis.
COXSTITUTIOX (IF THE ASSOCIATION iff OFFlflAL ACxRIfllTrRAl CHEfflSTS.
(1) This association shall be kno^wn as the Association of Official Agricultural
Chemists of the United States. The objects of the association shall be (1) to secure
uniformity and accuracy in the methods, results, and modes of statement of analysis
of fertilizers, soils, cattle, foods, dairy products, and other materials connected with
agricultiu-al industr>\; (2) to afford opportunity for the discussion of matters of interest
to agricultiu-al chemists.
(2) Anahtical chemists connected ^rith the United States Department of Agricul-
ture, or Trith any State or national agricultiu-al experiment station or agricultiu-al
college, or with any State or national institution or body charged with official con-
trol of the materials named in section 1. shall alone be eligible to membership: and
one such representative for each of these institutions or boards, when properly accred-
ited, shall be entitled to enter motions or vote in the association. Only such chemists
as are connected with institutions exercising official fertilizer control shall vote on
questions involving methods of analyzing fertilizers. All persons eligible to member-
ship shall become members ex officio and shall be allowed the privileges of mem-
bership at any meeting of the association after presenting proper credentials. All
members of the association who lose their right to such membership by retiring from
positions indicated as requisite for membership shall be entitled to become honorary
members and to have all privileges of membership save the right to hold office and
vote. All anah-tical chemists and others interested in the objects of the association
may attend its meetings and take part in its discussions, but shall not be entitled to
enter motions or vote.
(3) The officers of the association shall consist of a president, a vice-president, and
a secretary', who shall also act as treastuer; and these officers, together with two other
members to be elected by the a.ssociation. shall constitute the executive committee.
When any officer ceases to be a member by reason of withdrawing from a department
or board whose members are eligible to membership, his office shall be considered
vacant, and a successor may be appointed by the executive committee, to continue
in office till the anntial meeting next following.
(4 i There shall be appointed by the executive committee, at the regular annual
meeting, a referee and such associate referees for each of the subjects to be consid-
ered by the association as that committee may aeem appropriate.
It shall be the dut\' of these referees to prepare and distribute samples and stand-
ard reagents to members of the association and others desiring the same, to furnish
blanks for tabulating analyses, and to present at the annual meeting the results of
work done, discussion thereof, and recommendations of methods to be followed.
(5) The special duties of the officers of the association shall be fiuther defined,
when necessar}', by the executive committee.
(6) The annual meeting of this association shall be held at such place as shall be
decided by the association, and at such time as shall be decided by the executive
committee, and announced at least three months before the time of meeting.
(204)
205
(7) No changes shall be made in the methods of analysis used in official inspection,
except by unanimous consent, until an opportunity shall have been given all official
chemists having charge of the particular inspection affected to test the proposed
changes.
(8) Special meetings shall be called by the executive committee when in its judg-
ment it shall be necessary, or on the written request of five members; and at any
meeting, regular or special, seven enrolled members entitled to vote shall constitute
a quorum for the transaction of business.
(9) The executive committee will confer with the official boards represented with
reference to the payment of expenses connected with the meetings and publication
of the proceedings of the association.
(10) All proposed alternations or amendments to this constitution shall be referred
to a select committee of three at a regular meeting, and after report from such com-
mittee may be adopted by the approval of two-thirds of the members present entitled
to vote.
NDEX
Pa^e.
Acid, caffetannic, constitution, behavior toward reagents, methods of estima-
tion, etc 41-45
malic, test for obtaining value in maple products, results, etc 14-17
phosphoric, report by W. B. Kilgore, referee 157
salicylic, determinations . . ." 51-57
Adulteration, commercial fertilizers 177-178
Albrech, M. C. , analyst, maple sugar, comments 17
Alcohol, effect on colors 56
Alkaloids, determination in ipecac, cinchona and nux vomica 129-131
Allen, W. M,, report as associate referee on baking powders 28-29
Alumina, determination in phosphates 157-161
Ammonium salts, occurrence of chlorids 186
Analysis, methods, report of committee on revision, recommendations 148-150
Analysts, citral in flavoring extracts, list , 28
fats and oils, cloud and cold test, list, comments, etc 31-37
maple sugar products, list 16-17
moisture determination in foods, list 62-63
nitrogen determination, commercial meat extracts, comments 93-94
list, comments 78-83
potash in soils and mixed fertilizers, list, comments, etc 192-194
reducing sugars in sugar, massecuite, and molasses, list 119
salicylic acid, determination in tomatoes, list 57
water solution, list 52
wine, list 53
soil, list, comments 143-145
sucrose by optical methods, lists 122
sugar in condensed milk, comments 100-101
whiskies, cooperative work, list, comments. 21-23
Analytical results, unification of terms, committee 203
for reporting, report of committee 200
Arsenic, occurrence in chemical reagents 181-182
Ash constituents of plants, report, discussion 151-153
test for value in maple products, comments, etc 15-19
Assistant Secretary of Agriculture, address 161
Averitt, S. D., analyst, soils, comments 143, 145
Bailey, E. M., determination of malic acid in maple products, comments 17, 18
Baking powder, report by W. M. Allen, associate referee 28-29
Barium salts, occurrence of chlorids 187
Barnard, H . E. , report on beer analysis, reference 20
Beer analysis, report by H. E. Barnard, reference 20
recommendations 74
salicylic acid determination 52
Bigelow, W. D. , fruit products, comments 19-20
Boiling, George E., analyst, whiskies, comments 21, 22
BoswoRTH, Alfred W., paper on determination of the acidity of cheese 110-112
Bowker, W. H., determination of potash in soil and mixecl fertilizer, discus-
sion of methods 195-196
Boyle, M., and G. E. Patrick, subreport on dairy products 106-109
Bradley, H. C, analyst, separation of meat proteids, comments 93
Browne, C. A., jr., referee, and J. E. Halligan, associate referee, report on
sugar 1 16-1 25
(207)
208
Page.
Caffetannate, lead, solubility 43
Caff e tannic acid, constitution, behavior toward reagents 41-45
estimation, methods , 43-45
Calcium salts, occurrence of chlorids. 187
Caramel in vinegar, fuller's earth test, paper by W. L. Dubois 23-25
Carbon dioxid, determination in baking powders 28-29
Catsups, salicylic acid determination 55
Cereal products, report by A. McGill, associate referee 66-73
Chace, E. M., report as associate referee on flavoring extracts 25-28
Charron, A. T. , analyst, maple sugar, comments 17
Cheese, acidity, determination, by Alfred W. Bosworth 110-112
determinations, discussion 111-112
fat determinations by Gottlieb and ether extraction methods 109-110
report on separation of nitrogenous bodies, by R. Harcourt, referee . . 86-88
separation of nitrogenous bodies, instructions 86-87
Chemical reagents, testing, report of committee 181-188
Chemistry, agricultural, address by Dr. Alexius de 'Sigmond 188-189
Chlorids, occurrence in chemical reagents 182-188
Cinchona, determination of alkaloids 129-130, 134-142
Citral, determination in flavoring extracts, methods ' 25-28
Citrate solution, report by J. M. McCandless, associate referee 157-161
Cloud test, fats and oils, methods 30-31
Coffee and tea, recommendations 75
report by C. D. Howard, associate referee 41-45
tanninless coffee and coffee chaff, comparison of analyses 41
Colby, G. E., determination of kerosene in kerosene emulsion 165
Cold test, fats and oils 30-31
Coloring matters, experimental work, report by H. M. Loomis, reference .... 12
Colors, recommendations 1 2, 73
report by E. F. Ladd, associate referee 11-12
vegetable, detection, methods, subreport by A. G. Nickles 12-14
Committees, appointment 91
association, list 201-203
Condensed milk. See Milk.
Condiments (other than spices), analysis, methods 39-41
report by R. E. Doolittle, associate referee... 39-41
Constitution, Association Official Agricultural Chemists 204-205
Cook, F. C, report as associate referee on separation of meat proteids 91-98
Copper, reduced, comparison of methods for estimating 120-121
Crampton, C. a., report as associate referee on distilled liquors 20-23
vinegar test, comments 24
Cushman, Mr., determination of potash in soil and mixed fertilizer, discussion
of methods 194-195
Dairy products, recommendations 154
report by Albert E. Leach, referee 37-38
F. W. Woll, referee 98-99
subreport by G. E. Patrick and M. Boyle 106-109
State, and Food Officials, National Association, cooperation with com-
mittee on food standards, remarks of H. W. Wiley 173
Davidson, R. J., report as chairman of committee A on recommendations of
referees 196-199
on unification of terms. . . 200
Dickinson, W. E., analyst, determination of potash in soil and mixed fertilizer,
comments. 193
Distilled liquors. See Liquors.
Donk, M. G., analyst, determination of potash in soil and mixed fertilizer,
comments — 193
Doolittle, R. E., and F. O. Woodruff, tea analysis, methods review 46-50
report as associate referee on condiments (other than spices) . 39-41
Drugs and medicinal plants, recommendations 154
report by L. P. Kebler, referee 1 27-1 42
Dubois, W. L., article on fuller's earth test for caramel in vinegar 23-25
report as associate referee on food preservatives 51-58
Executive committee, association, additional members 201
Extracts, fiavoring, report by E. M. Chace, associate referee 25-28
209
Page.
Fat detenuiiiatioii in clieese by Gottlieb, and ether extraction metliodn 109-110
condensed milk 101-103
milk 106-109
recommendations 1 54
Fats and oils, report by L. M. Tolman, associate referee 29-37
Feeding stuffs and foods, recommendations 154
report by J. K. Haywood, referee 112-11.6
Feldspar, use as source of potash 194-196
Fertilizer legislation, committee 202
discussion 175-176
report of committee, by H. W. Wiley, chairman 174-176
Fertilizers, commercial, adulteration, paper by J. M. McCandless 177-178
detection of peat, paper by John Phillips Street 83-85
mixed, determination of potash, discussion, recommendations. . 190-196
Flavoring extracts, report by E. M. Chace, associate referee 25-28
Flour, expansion, determination, results 70
foreign starches, methods of detection 72-73
iodin value before and after bleaching 69
separation of vegetable proteids, report by Harry Snvder, associate
referee ". 88-90
Fkurs, classification 68-71
Food, adulteration, note 11
plant, definition, report of committee, by H. W. Wiley, chairman... 178-180
preservatives, recommendations 75
report by W. L. Dubois, associate referee 51-58
standards, committee 202
1906, report of committee , 168-173
Foods and feeding stuffs, recommendations 154
report, by J. K. Haywood, referee 112-116
water determination, rep^ort by F. C. Vv'eber, associate referee 58-65
Fkear, William F., determination of potash in soil and mixed fertilizers, dis-
cussion of methods 195
report as chairman of committee on food standards 168-173
Fruit products, report by Hermann C. Lythgoe, associate referee 19-20
Fuller's earth test for caramel in vinegar, paper by W. L. Dubois 23-25
Fusel oil, publications 23
GiBBOXEY, James H., nitrogen work, comments, recommendations 81-83
report as referee on nitrogen 76-83
Gladding, Thomas S., analyst, determination of potash in soil and mixed fer-
tilizer, comments 192
article on nitrogen estimations 85
Gliadin, separation from flour, methods 72
Gluten, crude and true, analyses 67-68
estimation in cereal products, conditions affecting 66-67
relation to total proteids in flours 68
Gottlieb method of analyzing milk 104-105
Graves, J. E., analyst, nitrogen work, comments 80
Grindley, H. S., analyst, separation of meat proteids, comments 93
Gudeman, Edward, analyst, whiskies, comments 21, 22
Gunning method of nitrogen estimations in fertilizers, comparison with
Kjeldahl method .' 85
Halligan, J. E., associate referee, and C. A. Browne, jr., referee, report on
sugar 116-125
Harcourt, R., report as referee on the separation of nitrogenous bodies 86-88
Hartwell, B. L., determination of potash in soil and mixed fertilizer, discus-
sion of methods 195
Hays, Willet M., Assistant Secretary of Agriculture, address 161
Haywood, J. K., paper on methods of analysis of lead arsenate 165-168
report as referee on foods and feeding stuffs 112-116
Heimburger, L., analyst, determination of potash in soil and mixed fertilizer,
comments 193
Holland, E. B., and P. H. Smith, analysts, nitrogen work, comments 81
chairman committee B on recommendations of referees, re-
port 154-157
31104— Xo. 105—07 14
210
Pa^e.
Hopkins, Cvril G., president of association, address, reference • 91
HoRNE, W. D., article on determination of suliDhites in sugar products 125-126
Howard, C. D. , report as associate referee on tea and coffee 41-45
Infants' and invalids' foods, recommendations 74
Insecticides, investigations, recommendations 199
Ipecac, determination of alkaloids 130
methods, results, remarks 1 86-139
Iron, determination in phosphates, report 157-161
Jackson, Holmes C, analyst, separation of meat proteids, comments 94
Jams, salicylic acids, determination 57, 58
Job, Robert, analyst, fats and oils, comments 34
Jones, C. H., maple-sugar products, report, comments, recommendations, etc. 15-19
report as associate referee on saccharine products 14-19
Kebler, L. F., remarks on drugs 141, 142
report as chairman on the testing of chemical reagents 181-188
referee on medicinal plants and drugs 127-142
Kerosene emulsion, determination of kerosene by centrifugal method 165
KiLGORE, B. VV., report as referee on phosphoric acid 157
Kjeldahl method of nitrogen estimations in fertilizers, comparison with Gun-
ning method 85
Xnisely, a. . L. , report as referee on potash 190-196
Ladd, E. F. , report as associate referee on colors 11-12
Lasche, A., analyst, whiskies, comments 21, 22
Law, L. M. , analyst, whiskies, comments 21 , 22
Leach, Albert E. ,_ report as referee on dairy products 37-38
Lead arsenate, analyses, results 166-168
analysis, methods, paper by J. K. Haywood 165-168
caffetannate, solubility 43
Legislation, ferti lizer, report of committee 174-176
J^iquors, distilled, recommendations 74
report by C. A. Crampton, associate referee 20-23
LooMis, H. M., analyst, whiskies, comments 21-22
report on experimental work in coloring matters, reference.. 12
Lowenstein, A., analyst, fats and oils, comments 35-36
Lythgoe, Hermann C., report as associate referee on fruit products 19-20
chairman of committee C on recommenda-
tions of referees 73-76
Magnesium salts, occurrence of chlorids ' 1 87
Maize starch, determination in wheat flour 72-73
Manns, A. G. , analyst, fats and oils, comments 31-33
Maple sirup, report by C. H. Jones, associate referee 14-19
sugar, report by C. H. Jones, associate referee 14-19
Marck worth, O. S., analyst, whiskies, comments 21, 22
Marmalades, salicylic acid, determination 57-58
Massecuite, sugar and molasses, determinations of total solids 117-119
McCandless, J. M., adulteration of commercial fertilizers, discussion 177-178
report as associate referee on determination of iron and
alumina in phosphates and on the citrate solution. .. 157-161
McDonnell, C. C. , analyst, nitrogen work, comments 80
McGiLL, A., report as associate referee on cereal products 66-73
Meat analysis, directions 91-92
proteids, separation, report by F. C. Cook, associate referee 91-98
Medicinal plants and drugs, recommendations 154
report, by L. F. Kebler 127-142
Members and visitors present, list 7-11
Mendel, L. B. , analyst, separation of meat proteids, comments 94
Methods, revision, committee, list. . .' 203
Milk, analysis, Gottlieb method, results, recommendations 104-106
condensed, determination of fats 101-103
sugar, analyses 100
dried, determination of fat 104
211
Paga
Milk, fat determinations - 10(>-109
powders, determination of fat 104
Miller, J. A., analyst, maple sugar, comments 17
Millwood, J. P., analyst, fats and oils, comments 33
Molasses, sugar, and massecuite, determinations of total solids 117-119
work, methods, recommendations 155-156
Morgan, Jerome J. , analyst, nitrogen work, comments 80
Morphine in opium, determination, methods 128-129, 132-134
Mory, A. V. H., analyst, fats and oils, comments 33
MuNsoN, L. S., report as associate referee on chemical methods of sugar
analysis 125
NicKLEs, A. G. , subreport on detection of vegetable colors 12-14
Nitrogen, determination in cheese, instructions, work, recommendations 86-87
estimations, comparison of Kjeldahl and Gunning methods, report
by Thomas S. Gladding 85
report by James H. Gibboney, referee 76-83
work, 1906, instructions 76-77
recommendations 196-197
Nitrogenous bodies in cheese, separation, report by K. Harcourt, referee 86-88
separation, methods, recommendations 197-198
Nomenclature, analytical results, committee on unification 203
report 200
Nominations, committee, report by C. D. Woods, chairman 150
Norton, J. H. , analyst, nitrogen work, comments 80
Nux vomica, determination of alkaloids 130-131
methods and results 136-137, 139-142
Ofiicers, association, list 201
Oil, fusel, literature 23^
Oilar, E. D. , analyst, fats and oils, comments 34-35
Oils and fats, report by L. M. Tolman, associate referee 29-37
Olson, Geo. A., and F. W. WoU, analysts, nitrogen work, comments 81
paper on fat determinations in cheese by the Gottlieb and the
ether-extraction methods 109-1 10
Opium, determination of morphine, cooperative work 132-134
powdered, determination of morphine 128-129
Patrick, G. PI, and M. Boyle, subreport on dairy products 106-109
Patten, Andrew J., comments on soil analysis 145
Peat, analyses, seven varieties 84
Pentosans, determinations in foods 115-116
Pettit, J. H., and A. Ystgard, determination of total potassium in soils... 147-148
report as referee on soils 142-147
Phosphates, determination of iron and alumina, and the citrate solution, report
by J. M. McCandless 157-161
methods, discussion 157-161
Phosphoric acid, report Dy B. W. Kilgore, referee 157
Phosphorus, determination in soils, methods 145-147
separation from meat 97
Plant constituents, inorganic, determination, recommendations 198,
report by W. W. Skinner, referee 151-153
food, definition, report of committee 178-180
Plants, medicinal. Hve Medicinal plants.
Pope, W. B. , analyst, maple sugar, comments 17
Potash, determination in mixed fertilizers 191
soil and fertilizer, recommendations 198
mixed fertilizers, comparison of official and
• modified methods 192
methods 190-191
Potash, feldspar as source, remarks 194, 196
report by A. L. Knisely, referee 190-196
Potassium, determination in soils 147-148
salts, C)ccurrence of chlorids 184-1 85
President, Association, committee on address 203
report of committee 199-200
212
Page.
Proteids, meat, separation, methods, recommendations 198
report by F. C. Cook, associate referee 91-98
milk and cheese, separation, recommendations 197
vegetable, separation from wheat and flour, report by Harry Snyder,
associate referee 88-90
recommendations 197-198
Reagents, chemical, testing committee 202
report of committee, L. F. Kebler, chairman . . 181-188
Recommendations 23, 28, 45, 57-58, 65, 87-88, 105-106
analysis of maple sugar products 19
determination of inorganic plant constituents 152-153
of committee on revision of methods of analysis 149
referees, report of committee A, R. J. Davidson, chair-
man 196-199
B, E. B. Holland, chair-
man 154-157
C, H. G. Lythgoe, chair-
man 73-76
soil analyses '----. - - 147
Reed, H. C, report as referee on tannin 161-164
Referees, associate, association, list 201-202
association, list 201
Resolutions, report of committee 200
Richardson, W. D. , analyst, fats and oils, comments 33
Rosenstein, L., analyst, nitrogen work, comments '. 80
Saccharine i^roducts (maple sugar and sirup), report by C. H.Jones, associate
referee 14-19
recommendations 74
Salicylic acid, determinations 51-57
Schmidt, A. H., analyst, fats and oils, comments 35
Shieb, S. H., analyst, nitrogen work, comments 80
'Sigmond, Dr. Alexius de, address 188-190
Sirup, maple and sugar, report by C. H. Jones, associate referee 14-19
Skinner, W. W., report as referee on inorganic plant constituents 151-153
Smith, E. E., analyst, separation of meat proteids, comments 93
Smith, P. H., and E. B. Holland, analysts, nitrogen work, comments 81
Snyder, Harry, comments on soil analysis 145
report as associate referee on the separation of vegetable pro-
teids from wheat and flour 88-90
Sodium salts, occurrence of chlorids 185-186
Soils, analyses, methods, results „ 143-147
analysis, directions 142-143
methods, recommendations 198-199
recommendations 147
• determination of potash, discussion, recommendations 190-196
total phosphorus, methods 145-147
report by J. H. Pettit, referee 142-147
Stallings, R. E., analyst, whiskies, comments 21 , 22
Standards, food, 1906, report of committee 168-173
Starches, foreign, determination in wheat flour 72-73
Stookey, L. B., analyst, separation of meat proteids, comments 94
Street, John Phillips, analyst, nitrogen work, comments 80
article on detection of peat in commercial fertilizers. . 83-85
Sucrose, determination by chemical methods 124-125
optical methods 121-124
Sugar analysis, chemical methods, report by L. S. Munson, associate referee.. 125
recommendations 154-156
maple, and sirup, report by C. H. Jones, associate referee 14-19
massecuite and molasses, determinations of total solids 117-119
products, chemical determination of sulphites 125-127
recommendations .~ 154-156
report by C. A. Browne, jr., referee, and J. E. Halligan, associate
referee ^. 116-125
Sugars, reducing, determination 119-121
213
Page.
Sulphites in sugar ])ioducts, determination 125-127
Sy, A. P. , analyst, maple sugar, comments 17
Tannin, cooperative work, etc., (U.scussion 162-164
determination in tea 48-49
recommendations 156
report by II. C Reed, referee 161-164
Tea, analyses, varieties examined 46
analysis, methods, review by II. E. Doolittle and F. O. Woodruff 46-50
and coffee, recommendations 75
report by C D. Howard, associate referee 41-45
drying, comparison of different methods 50
, moisture, determination j 50
water extract, determination, methods 46-48
Terms for reporting analytical results, unification, committee 203.
report of committee 200
Thatcher, R. W., analyst, determination of potash in soil and mixed fertilizer,
conunents 192
Theine, determination in tea 49-50
ToLMAN, L. M. , analyses, fats and oils, comments 36-37
report as associate referee on fats and oils 29-37
Tomatoes, salicylic acid, determination, results 57 "
Trescot, T. C. , analyst, nitrogen work, comments 81
Valin, A., analyst, maple sugar, comments 17
Van Slyke, L. L. , report of committee on resolutions 200
Vegetable proteids, separation, report by Harry Snyder 88-90
Vinegar, fuller's earth test for caramel 23-25
Water in foods, determination, recommendations 75
Weber, F. C, report as associate referee on determination of water in foods. . 58-65
J. E. , analyst, fats and oils, comments 33
Welton, F. A., analyst, determination of potash in soil and mixed fertilizer,
comments , 193
Wheat, vegetable proteids, separation, report by Harry Snyder, associate
referee 88-90
Whisky, artificial, analyses 21
bourbon, analyses 21
rye, analyses 21
See also Liquors.
AVickhorst, Max H., analyst, fats and oils, comments 33
Wiley, H. W., remarks on drugs, ash, inorganic constituents, and sulphids.. 126,
141, 153
food standards committees 173
revision of methods of analysis 148-149
report as chairman committee on fertilizer legislation 174-176
Wine, salicylic acid determination 52-55
Woodruff, F. 0., and R. E. Doolittle, tea analysis, methods, review 46-50
Woods, C. D. , report as chairman committee on nominations 150
WoLL, F. W., and Geo. A. Olsen, analysts, nitrogen work, comments 81
report as referee on dairy products 98-99
YsT(4ARD, A., and J. H. Pettit, paper on determination of total potassium in
soils 147-148
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11